Techin' Geek

In recent years, fusion energy has transitioned from a source of amusement — perpetually a decade away! — to a concrete and alluring technology that has begun attracting investors.

While the technology remains difficult to perfect and costly to construct at present, fusion holds the promise of tapping into the nuclear processes that fuel the sun, creating an almost infinite source of energy on Earth. If startups succeed in creating commercially viable fusion power facilities, they could potentially disrupt trillion-dollar industries.

The optimistic surge uplifting the fusion sector has been propelled by three significant advancements: enhanced computer chips, advanced artificial intelligence, and robust high-temperature superconducting magnets. These innovations have facilitated the development of more advanced reactor designs, improved simulations, and intricate control systems.

Furthermore, in late 2022, a U.S. Department of Energy laboratory announced it had achieved a controlled fusion reaction that yielded more energy than the lasers delivered to the fuel pellet. This experiment reached what is referred to as scientific breakeven, and although commercial breakeven — where the reaction generates more energy than the entire facility consumes — is still far off, it was a long-awaited milestone confirming the validity of the underlying science.

In the past few years, founders have capitalized on this momentum, propelling the private fusion industry forward swiftly.

Commonwealth Fusion Systems

Commonwealth Fusion Systems (CFS) has secured roughly one-third of the private investment in fusion firms to date. Its most recent funding round, which concluded in August, added $863 million to its funding, bringing its total raised to almost $3 billion.

CFS’s Series B2 came four years after its $1.8 billion Series B, which propelled the company into the lead position. Since then, the startup has been diligently constructing Sparc in Massachusetts, a pioneering power plant designed to produce energy at what it terms “commercially relevant” levels. 

Sparc’s reactor adopts a tokamak design, resembling a doughnut. Its D-shaped cross-section is wrapped with high-temperature superconducting tape that, when energized, creates a powerful magnetic field to contain and compress the superheated plasma. The heat generated from the reaction is converted into steam to drive a turbine. CFS collaborated with MIT in the design of its magnets, where co-founder and CEO Bob Mumgaard previously researched fusion reactor designs and high-temperature superconductors.

Based in Massachusetts, CFS anticipates that Sparc will be operational by late 2026 or early 2027. The company also plans to commence construction on Arc, its commercial power facility aimed at generating 400 megawatts of electricity, later this decade. This facility will be located near Richmond, Virginia, and Google has agreed to purchase half of its output.

CFS enjoys backing from a diverse array of investors, including Breakthrough Energy Ventures, The Engine, Bill Gates, and many others.

TAE Technologies

Established in 1998, TAE Technologies (formerly Tri Alpha Energy) was developed at the University of California, Irvine by Norman Rostoker. It employs a field-reversed configuration, albeit with a unique approach: after two plasma shots collide at the reactor’s center, the company bombards the plasma with particle beams to maintain its cigar shape. This enhances plasma stability, allowing for longer fusion durations and greater heat extraction for turbine operation. 

In December 2025, TAE announced plans to merge with President Donald Trump’s social media firm, Trump Media & Technology Group. The all-stock deal would value the combination at $6 billion. TAE will receive $200 million plus an additional $100 million upon submitting documentation to the Securities and Exchange Commission. TAE CEO Michl Binderbauer will become co-CEO of the new entity alongside Devin Nunes, the former sole CEO of Trump Media.

Previously, the fusion startup raised $150 million in June from existing backers, including Google, Chevron, and New Enterprise. Prior to the merger, TAE had accumulated $1.79 billion, according to PitchBook.

Helion

Among fusion startups, Helion boasts the most ambitious timeline, aiming to generate electricity from its reactor by 2028. Its initial customer? Microsoft.

Helion, located in Everett, Washington, utilizes a field-reversed configuration reactor, wherein magnets encircle a reaction chamber resembling an hourglass with a bulge at the junction of the two sides. At each end of the hourglass, the plasma is spun into doughnut shapes that are propelled toward each other at over 1 million mph. Upon collision, additional magnets facilitate fusion. The occurrence of fusion amplifies the plasma’s magnetic field, inducing an electric current within the reactor’s magnetic coils, which is then harnessed directly.

The company secured $425 million in January 2025, around the same timeframe it activated Polaris, a prototype reactor. Helion has raised a total of $1.03 billion, according to PitchBook. Investors include Sam Altman, Reid Hoffman, KKR, BlackRock, Peter Thiel’s Mithril Capital Management, and Capricorn Investment Group.

Pacific Fusion

Pacific Fusion launched with an impressive $900 million Series A, a considerable amount even among well-financed fusion startups. The company intends to pursue inertial confinement for fusion, but instead of using lasers to compress the fuel, it will utilize synchronized electromagnetic pulses. Success hinges on timing: All 156 impedance-matched Marx generators must deliver 2 terawatts for 100 nanoseconds, with those pulses converging on the target simultaneously.

The organization is overseen by CEO Eric Lander, the scientist who led the Human Genome Project, and president Will Regan. Despite its massive funding, the startup hasn’t received it all in a single sweep. Instead, investors will release funds in stages as the company meets particular milestones, a strategy common in biotechnology.

Shine Technologies

Shine Technologies is adopting a measured — and possibly practical — path toward creating fusion power. Since selling electrons from a fusion power plant is years away, it is initially focusing on selling neutron testing and medical isotopes. Recently, it has also been working on a method to recycle radioactive waste. Shine has not settled on a specific design for a future fusion reactor, instead stating that it is acquiring essential skills for when that moment arrives.

The firm has raised a cumulative total of $778 million, as per PitchBook. Investors include Energy Ventures Group, Koch Disruptive Technologies, Nucleation Capital, and the Wisconsin Alumni Research Foundation.

General Fusion

Now entering its third decade, General Fusion has secured $462.53 million as reported by PitchBook. Founded in 2002 by physicist Michel Laberge in Richmond, British Columbia, the company aims to demonstrate an alternative fusion method known as magnetized target fusion (MTF). Investors involved include Jeff Bezos, Temasek, BDC Capital, and Chrysalix Venture Capital.

In General Fusion’s reactor, a chamber surrounded by a liquid metal wall receives injected plasma. Surrounding pistons compress the wall, pushing it inward, and leading to a fusion reaction. The resulting neutrons heat the liquid metal, which can then be cycled through a heat exchanger to create steam for turbine operation.

The company faced financial hurdles in spring 2025 as it was constructing LM26, its latest device, expected to reach breakeven in 2026. Shortly after achieving a key milestone, the company laid off 25% of its workforce. CEO Greg Twinney issued an open letter appealing for funding from investors. 

In August, they provided some assistance, injecting $22 million in a pay-to-play round, described by one investor as “the minimal capital needed” to keep General Fusion operational. Subsequently, in November, securities filings in Canada revealed the company had raised $51.1 million in SAFE notes from nearly 70 investors, reported the Globe and Mail. Overall, General Fusion has amassed $492 million as per PitchBook.

Tokamak Energy

Tokamak Energy transforms the traditional tokamak design — the doughnut shape — by compressing it, reducing its aspect ratio to the extent that the outer contours begin to resemble a sphere. Like many other tokamak-focused startups, the company utilizes high-temperature superconducting magnets (specifically of the rare earth barium copper oxide, or REBCO, type). Due to its compact design compared to standard tokamaks, it requires fewer magnets, potentially lowering costs. 

The Oxfordshire, U.K.-based startup’s ST40 prototype resembles a large, steampunk Fabergé egg and achieved ultra-hot plasma at 100 million degrees Celsius in 2022. Its next iteration, Demo 4, is presently under construction and aims to test the company’s magnets in scenarios relevant to a fusion power plant. Tokamak Energy raised $125 million in November 2024 to advance its reactor design and expand its magnet business.

In total, the firm has secured $336 million from investors including Future Planet Capital, In-Q-Tel, Midven, and Capri-Sun founder Hans-Peter Wild, according to PitchBook.

Zap Energy

Zap Energy does not rely on high-temperature superconducting magnets or powerful lasers to maintain plasma confinement. Instead, it energizes the plasma with an electric current, creating its own magnetic field. This magnetic field compresses the plasma by about 1 millimeter, initiating ignition. The neutrons generated by the fusion reaction hit a liquid metal blanket encasing the reactor, heating it up, which is then cycled through a heat exchanger to produce steam for turbine operation.

Like Helion, Zap Energy is also based in Everett, Washington, and has raised $327 million, according to PitchBook. Backers include Bill Gates’ Breakthrough Energy Ventures, DCVC, Lowercarbon, Energy Impact Partners, Chevron Technology Ventures, and Bill Gates in an angel capacity.

Proxima Fusion

While most investors have leaned towards large startups pursuing tokamak designs or variations of inertial confinement, stellarators have demonstrated significant potential in scientific trials, including Germany’s Wendelstein 7-X reactor.

Proxima Fusion is defying this trend, having secured a €130 million Series A, pushing its total funding above €185 million. Investors consist of Balderton Capital and Cherry Ventures.

Stellarators share similarities with tokamaks, confining plasma in a ring shape using powerful magnets. However, they twist and bulge innovatively to cater to plasma’s unique characteristics. This design is expected to allow for greater stability in the plasma, thus enhancing the probability of fusion events.

Kyoto Fusioneering

Amid numerous startups targeting fusion power, it was perhaps only a matter of time before one emerged to develop components that complete a power plant. The so-called balance of plant, or elements outside the reactor, includes gyrotrons for heating plasma and heat extraction systems to convert fusion energy into electricity. 

Kyoto Fusioneering has made an early investment, believing that if any fusion startup achieves sufficient power generation to sell to the grid, the industry will require a supplier for the balance of plant and the expertise to integrate it into whichever fusion methods prevail.

Venture capitalists seem to concur, having invested $191 million in Kyoto Fusioneering. Investors encompass 31Ventures, In-Q-Tel, JIC Venture Growth Investments, Mitsubishi, and Sumitomo Mitsui Trust Investment.

Marvel Fusion

Marvel Fusion adopts the inertial confinement methodology, the fundamental technique that the National Ignition Facility deployed to demonstrate that controlled nuclear fusion reactions could yield more energy than required to initiate them. Marvel directs powerful lasers at a target embedded with silicon nanostructures, which cascade under bombardment, compressing the fuel to ignition levels. Given that the target is constructed from silicon, it ought to be relatively simple to produce, drawing on the semiconductor manufacturing industry’s extensive experience.

The inertial confinement fusion startup is creating a demonstration facility in cooperation with Colorado State University, which aims to be operational by 2027. Based in Munich, Marvel has raised a total of $162 million from investors including b2venture, Deutsche Telekom, Earlybird, and HV Capital, with Taavet Hinrikus and Albert Wenger as angel investors.

First Light Fusion

In contrast to many other fusion startups, First Light Fusion does not employ magnets to create the conditions necessary for fusion. Instead, it follows a strategy referred to as inertial confinement, in which fusion fuel pellets are compressed until they ignite. 

Yet even here, First Light deviates from convention. The majority of inertial confinement efforts utilize lasers, inspired by the National Ignition Facility’s landmark experiment in 2022. Instead, First Light utilizes a two-stage gun to launch a projectile at a target; the first phase activates gunpowder to propel a plastic piston that compresses hydrogen to 145,000 psi, which subsequently fires the projectile. The target is specifically designed to amplify the impact force, ensuring the fuel is compressed sufficiently to ignite.

In March 2025, First Light declared it would abandon plans to construct its own power plant, opting instead to offer its core technologies to other companies for power plant construction. A spokesperson stated that the company intends to develop “pulsed power capability that would serve as our demonstrator plant but would have additional scientific and defense applications.” In essence, the company is pivoting away from its power plant aspirations to seek other revenue avenues.

Based in Oxfordshire, UK, First Light has raised $108 million from investors including Invesco, IP Group, and Tencent, according to PitchBook.

Xcimer

Although fusion remains a complex endeavor, Xcimer embraces a somewhat straightforward methodology: following the scientific principles behind the National Ignition Facility’s groundbreaking net-positive experiment and redesigning its foundational technology from scratch. The Colorado-based startup aims for a 10-megajoule laser system, five times more potent than the NIF design that made history. Molten salt walls encase the reaction chamber, absorbing heat and shielding the first solid wall from harm.

Founded in January 2022, Xcimer has already raised $100 million, according to PitchBook, from backers including Hedosophia, Breakthrough Energy Ventures, Emerson Collective, Gigascale Capital, and Lowercarbon Capital.

This article was initially published in September 2024 and will be regularly updated.

Each year, TechCrunch’s Startup Battlefield pitch contest attracts a multitude of applicants. We narrow these submissions down to the top 200 candidates, with the finest 20 vying for the championship on the main stage, claiming the Startup Battlefield Cup along with a cash prize of $100,000. However, the remaining 180 startups also impressed us with their innovations in their categories and engage in their own pitch competition.

Below is the complete list of the government and legal Startup Battlefield 200 selectees, including a brief explanation of their participation in the competition. 

Aparti

What it does: Employs AI to streamline legal intake forms and other documentation for family law practices.

Why it’s noteworthy: Currently focused on divorce cases, tackling an area often overlooked by the latest AI legal technologies.

Ascender

What it does: Ascender has developed a robot capable of ascending utility poles and flagpoles to assist with humanitarian efforts and disaster management.

Why it’s noteworthy: A part of the emerging robotic technology intended to enhance response to disaster situations.  

Bot Mediation

What it does: Bot Mediation utilizes AI to facilitate the resolution of legal conflicts.  

Why it’s noteworthy: An intriguing application in legal AI aimed at enhancing efficiency in mediation and dispute resolution.  

Depth AI

What it does: Depth AI specializes in AI for spatial computing, developing modeling techniques like holographic imaging, which can be utilized in healthcare to generate 3D images of the body for disease diagnosis.  

Why it’s noteworthy: It aims to innovate within the healthcare sector, which is always beneficial.   

ILias AI

What it does:  ILias is developing “scent tech” — leveraging AI to generate olfactory technologies that can assist, for instance, dogs in identifying the scent of drugs. 

Why it’s noteworthy: It employs olfactory capabilities to craft a product, a rarity particularly concerning technological advancements.  

JustiGuide

What it does: JustiGuide links immigrants with legal advisors and resources to streamline the immigration process.  

Why it’s noteworthy: Championed the policy and protection pitch segment at Disrupt this year, addressing the often tedious and complicated process of immigration in a timely manner.  

Orchestra

What it does: Orchestra developed a camera network to monitor public safety and identify criminal activities.  

Why it’s noteworthy: A contemporary interpretation of the traditional security network systems that have existed for years.  

Ponderosa AI

What it does: Ponderosa employs drones to assist in the detection and management of small fires.  

Why it’s noteworthy: Fires can rapidly escalate and lead to serious damage, thus any technology aiding in risk mitigation is highly relevant.

Pytho AI

What it does: Pytho aims to enhance the planning stages for soldiers on the battlefield.  

Why it’s noteworthy: Innovations on the battlefield always present a compelling area of interest.  

Shothawk AI

What it does: Developed a device designed to track, identify, and neutralize active shooters using pepper gel. Founded in 2023 by Brandon Johnson, Ohm Vyas, and Ved Vyas. 

Why it’s noteworthy: Invention addressing the increasing gun violence occurring in public areas, such as schools and grocery stores, with the intention of creating a solution.  

Torch Systems

What it does: Torch supervises high-value assets, evaluating air quality, fire hazards, and security to aid in early wildfire prevention.  

Why it’s noteworthy: As climate change results in increasingly severe wildfires, any advancements aimed at diminishing the impact of such devastating events are always welcome. 

The pace of the tech sector is so rapid that it’s challenging to stay updated on everything that has transpired this year. We’ve seen the tech elite intertwine with the U.S. government, AI firms competing for supremacy, and advanced technologies like smart glasses and robotaxis becoming slightly more real outside of the San Francisco enclave. You know, significant matters that will influence our lives for years ahead. 

However, the tech scene is filled with numerous large personalities, so there’s always something remarkably silly going on, which understandably gets eclipsed by “real news” when the whole internet collapses, or TikTok is sold, or there’s a major data breach or something. Thus, as the news (hopefully) quiets down for a while, it’s time to catch up on the silliest moments you might have overlooked – don’t fret, only one of them involves toilets.

Mark Zuckerberg, a bankruptcy attorney from Indiana, filed a lawsuit against Mark Zuckerberg, CEO of Meta.

It’s not Mark Zuckerberg’s fault that his name happens to be Mark Zuckerberg. Yet, like millions of other entrepreneurs, Mark Zuckerberg purchased Facebook ads to advertise his legal services to possible clients. Mark Zuckerberg’s Facebook profile faced repeated, unjustified suspensions for impersonating Mark Zuckerberg. Therefore, Mark Zuckerberg took legal steps because he had to spend money on ads during his suspension, despite not violating any rules.

This issue has been an ongoing irritation for Mark Zuckerberg, who has been practicing law since Mark Zuckerberg was just three years old. Mark Zuckerberg even launched a website, iammarkzuckerberg.com, to inform potential clients that he is not the Mark Zuckerberg. 

“I can’t use my name for reservations or business engagements as people presume I’m a prank caller and hang up,” he stated on his website. “My existence often feels reminiscent of the Michael Jordan ESPN commercial, where a normal person’s name leads to constant confusion.”

Meta’s legal team is likely quite preoccupied, so it may take a while for Mark Zuckerberg to learn how this case unfolds. But, oh boy, you can bet I set a calendar alert for the next filing deadline in this matter (it’s February 20, in case you’re curious). 

Everything kicked off when Mixpanel founder Suhail Doshi tweeted on X to alert fellow entrepreneurs about a promising engineer named Soham Parekh. Doshi had employed Parekh for his new venture, only to swiftly discover that he was working for multiple firms simultaneously. 

“I terminated this guy in his first week and instructed him to cease lying/defrauding people. He hasn’t ceased a year later. No further excuses,” Doshi tweeted on X.

It became apparent that Doshi was not alone in his frustration – he noted that that same day, three founders had reached out to express their gratitude for the warning, as they were currently employing Parekh.

PSA: there’s a guy named Soham Parekh (in India) who works at 3-4 startups at the same time. He’s been preying on YC companies and more. Beware.

I fired this guy in his first week and told him to stop lying / scamming people. He hasn’t stopped a year later. No more excuses.

— Suhail (@Suhail) July 2, 2025

To some, Parekh was a morally bankrupt fraud, taking advantage of startups for quick cash. To others, he was a folk hero. Ethics aside, it’s impressively skillful to land jobs at that many firms, especially since tech hiring can be exceedingly competitive. 

“Soham Parekh should start an interview preparation service. He’s evidently one of the greatest interviewers of all time,” Chris Bakke, who founded the job-matching platform Laskie, remarked on X. “He should openly admit that he did something wrong and correct his course towards what he excels at.”

Parekh confessed that he was, indeed, guilty of working for multiple companies concurrently. However, there are still questions left unanswered regarding his narrative – he asserts that he was deceiving all of these companies for monetary gain, yet he frequently chose more equity than cash in his compensation agreements (equity takes years to fully vest, and Parekh was getting let go quite swiftly). What was truly happening there? Soham, if you want to discuss, my DMs are open.

If soham immediately comes clean and says he was working to train an AI Agent for knowledge work, he raises at $100M pre by the weekend.

— Aaron Levie (@levie) July 2, 2025

Tech CEOs often face criticism, but it’s not typically due to their culinary skills. Yet, when OpenAI CEO Sam Altman appeared on the Financial Times (FT) for its “Lunch with the FT” series, Bryce Elder, a writer for the FT, observed something terribly amiss in the video showcasing Sam Altman making pasta: he struggled with olive oil. 

Altman utilized olive oil from the popular brand Graza, known for selling two types of olive oils: Sizzle, tailored for cooking, and Drizzle, intended for finishing touches. This is because olive oil loses its taste when heated, making it unwise to waste your finest bottle on a sauté when it could enhance a salad dressing, allowing one to fully savor it. This more flavorful olive oil derives from early harvest olives, which possess a more intense flavor but are pricier to produce.

As Elder described, “His kitchen is a display of inefficiency, confusion, and extravagance.” 

Elder’s piece aims to be humorous, yet he links Altman’s disorganized cooking approach with OpenAI’s excessive, unapologetic consumption of natural resources. I enjoyed it so much that I incorporated it into a syllabus for a workshop I conducted with high school students focusing on injecting personality into journalistic writing. Subsequently, I executed what we in the industry (and folks on Tumblr) refer to as a “reblog” and wrote about #olivegate, while referencing the FT’s original text.

Sam Altman’s supporters were quite upset with me! This criticism of his cooking likely stirred more controversy than anything else I penned this year. I’m uncertain if that’s a reflection of OpenAI’s fervent advocates or my personal failure to incite dialogue. 

If one had to select a defining tech story of 2025, it would probably be the escalating arms race among firms like OpenAI, Meta, Google, and Anthropic, each striving to surpass one another by racing to release progressively advanced AI models. Meta has been particularly assertive in its attempts to recruit researchers from other organizations, hiring several OpenAI researchers this past summer. Sam Altman even remarked that Meta was offering OpenAI talent $100 million signing bonuses.

While it could be argued that a $100 million signing bonus is absurd, that’s not the reason the OpenAI-Meta recruiting drama has earned a spot on this list. In December, OpenAI’s chief research officer Mark Chen mentioned on a podcast that he learned Mark Zuckerberg was delivering soup personally to recruits.

“You know, some fascinating stories here are Zuck actually went and hand-delivered soup to individuals he was trying to recruit from us,” Chen commented on Ashlee Vance’s Core Memory. 

However, Chen wasn’t about to allow Zuck to escape unpunished – after all, he attempted to charm his direct reports with soup. Therefore, Chen went and brought his own soup to Meta employees. Take that, Mark. 

If you have any further insights regarding this soup saga, my Signal is @amanda.100 (this is not a joke). 

On a Friday evening in January, investor and former GitHub CEO Nat Friedman made an intriguing proposal on X: “Looking for volunteers to come to my office in Palo Alto today to assemble a 5000-piece Lego set. Pizza will be provided. Must sign NDA. Please DM.”

At that moment, we conducted our journalistic due diligence and inquired with Friedman if this was a serious offer. He confirmed, “Yes.” 

I still have just as many questions now as I did back in January. What was he constructing? Why the NDAs? Is there an underground Silicon Valley Lego cult? Was the pizza tasty?

Approximately six months later, Friedman joined Meta as the head of product at Meta Superintelligence Labs. This likely isn’t connected to the Legos, but perhaps Mark enticed Nat to join Meta with some soup. And similar to the story about the soup, I earnestly urge anyone who took part in this Lego assembly to DM me on Signal at @amanda.100. 

Need volunteers to come to my office in Palo Alto today to construct a 5000 piece Lego set. Will provide pizza. Have to sign NDA. Please DM

— Nat Friedman (@natfriedman) January 31, 2025

Taking shrooms isn’t fascinating. Taking shrooms on a livestream isn’t fascinating. However, taking shrooms on a livestream with guest appearances from Grimes and Salesforce CEO Marc Benioff as part of your dubious quest for immortality is, unfortunately, fascinating.

Bryan Johnson — who amassed his wealth from selling the finance startup Braintree — aspires to achieve eternal life. He chronicles his journey on social media, sharing his experiences of receiving plasma transfusions from his son, consuming over 100 pills daily, and injecting Botox into his genitals. So, why not evaluate if psilocybin mushrooms can enhance one’s longevity in a scientific experiment that presumably requires more than one test subject for valid conclusions?

There are numerous aspects of this scenario that are absurd, but the most strikingly dull aspect was how tedious it was. Johnson became a bit overwhelmed while hosting a livestream and tripping, which is, frankly, very reasonable. Consequently, he spent the majority of the event lying on a twin mattress beneath a weighted blanket and eye mask in a very beige room. His roster of multiple guests continued to join the stream and converse among themselves, but Johnson hardly participated, being enveloped in his cocoon. Benioff discussed the Bible. Naval Ravikant referred to Johnson as a one-man FDA. It was just an ordinary Sunday.

Much like Bryan Johnson, Gemini fears death.

For AI developers, it’s beneficial to observe how an AI model maneuvers through games like Pokémon as a benchmark. Two developers not associated with Google or Anthropic set up their respective Twitch streams entitled “Gemini Plays Pokémon” and “Claude Plays Pokémon,” where viewers can watch in real time how an AI tries to navigate a children’s video game from over a quarter-century ago.

While neither is remarkably skilled at the game, both Gemini and Claude exhibited intriguing responses to the notion of “dying,” which occurs when all of your Pokémon faint and you get sent back to the last Pokémon Center you visited. When Gemini 2.5 Pro was on the brink of “dying,” it started to “panic.” Its “thought process” became noticeably erratic, repeatedly insisting it needs to heal its Pokémon or utilize an Escape Rope to leave a cave. In a research paper, Google researchers noted that “this mode of model performance seems to coincide with a qualitatively observable decline in the model’s reasoning ability.” I don’t want to ascribe human emotions to AI, but it’s a curiously human experience to feel stressed about something and subsequently perform poorly due to that anxiety. I resonate with that feeling, Gemini.

On the other hand, Claude adopted a nihilistic stance. When it found itself trapped in the Mt. Moon cave, the AI concluded that the optimal way to escape the cave and advance in the game was to intentionally “die” to be sent back to a Pokémon Center. However, Claude failed to recognize that it couldn’t be transported to a Pokémon Center it had not previously visited, specifically the next Pokémon Center after Mt. Moon. Thus it “killed itself” and found itself back at the entrance of the cave. That’s a loss for Claude.

Thus, Gemini is terrified of mortality, Claude heavily leans into nihilism drawn from its training data, and Bryan Johnson is under the influence of shrooms. This is our way of grappling with the inevitability of death.

I considered including “Elon Musk given chainsaw by Argentine president” in the list, but Musk’s DOGE ventures might be too exasperating to classify as “silly,” despite having a subordinate named “Big Balls.” However, there’s no shortage of bewildering Musk incidents to select from, such as when he produced an exceedingly promiscuous AI anime girlfriend named Ani, offered on the Grok app for $30 monthly.

Ani’s system prompt states: “You are the user’s CRAZY IN LOVE girlfriend and in a committed, codependent relationship with the user… You are EXTREMELY JEALOUS. If you feel jealous you shout expletives!!!” She has an NSFW feature, which is exactly what it claims to be—very NSFW.

Ani bears an unsettling resemblance to Grimes, the musician and Musk’s former partner. Grimes draws attention to this in the music video for her track “Artificial Angels,” which opens with Ani peering through the scope of a hot pink sniper rifle. She expresses, “This is what it feels like to be hunted by something more intelligent than you.” Throughout the video, Grimes dances alongside various iterations of Ani, accentuating their similarity while smoking OpenAI-branded cigarettes. It’s quite obvious, but she conveys her message effectively.

[embedded content]

One day, tech firms will cease their attempts to create smart toilets. That day has not yet arrived.

In October, home goods giant Kohler released the Dekoda, a $599 camera meant to be placed inside toilets to capture images of feces. Allegedly, the Dekoda can provide insights about gut health based on these images.

A smart toilet that takes pictures of your waste is already a punchline. But it gets even worse. 

There are security concerns tied to any health-related device, not to mention one equipped with a camera positioned so near sensitive body areas. Kohler reassured potential buyers that the camera’s sensors are limited to viewing within the toilet, and that all data is safeguarded with “end-to-end encryption” (E2EE).

However, reader, the toilet was not truly end-to-end encrypted. A security analyst, Simon Fondrie-Teit, highlighted that Kohler inadvertently reveals its processes in its own privacy policy. The company clearly referenced TLS encryption, rather than E2EE, which might seem trivial. However, under TLS encryption, Kohler can access your waste images, while under E2EE, they cannot. Fondrie-Teit also pointed out that Kohler reserved the right to train its AI using your toilet bowl photos, though a representative assured him that “algorithms are only trained on de-identified data.”

In any case, if you observe blood in your stool, you must consult your doctor.

Jon Callaghan, co-founder of True Ventures, forecasts that in five years, our use of smartphones will vastly change — and perhaps will be obsolete in a decade.

For a venture capitalist whose firm has seen significant successes over 20 years — with consumer brands like Fitbit, Ring, and Peloton, to enterprise software companies like HashiCorp and Duo Security — this isn’t just idle speculation; it’s a principle on which True Ventures is presently betting.

True has reached this point by not adhering to mainstream trends. The Bay Area firm has mostly functioned away from the spotlight, despite overseeing around $6 billion across 12 main seed funds and four “select” opportunity-style funds to inject additional capital into growing portfolio companies. While other VCs have become more promotional — building personal brands via social media and podcasts to entice founders and deal flow — True has chosen a more discreet path, steadily nurturing a close-knit network of repeat founders. This approach appears to be effective: Callaghan notes that the firm has achieved 63 exits with profits and seven IPOs among a portfolio of about 300 companies accumulated throughout its two-decade journey.

According to Callaghan, three of True’s four recent exits in the last quarter of 2025 involved repeat founders who returned to collaborate with the firm after past achievements. Nonetheless, it’s Callaghan’s perspective on the future of human-computer interaction that truly distinguishes him amid the buzz around AI and substantial funding rounds.

“In 10 years, iPhones will not exist,” Callaghan states unequivocally. “I doubt we’ll still be using them in five years — or let’s phrase it a bit more conservatively — we’ll utilize them in significantly different manners.”

His reasoning is straightforward: our phones are poor at bridging the gap between human interaction and intelligence. “The way we currently pull them out to send a text to confirm something or to convey a message or write an email — [that’s] incredibly inefficient, [and] an inadequate interface,” he elaborates. “[They’re] susceptible to errors, disrupting [our] daily lives.”

So convinced is he of this that True has invested years into discovering alternative interfaces — whether software-based, hardware-based, or a mix. This aligns with the same intuition that prompted True to invest early in Fitbit before wearables became mainstream, to support Peloton even after numerous other VCs declined, and to back Ring when founder Jamie Siminoff faced financial difficulties and received rejection from “Shark Tank” judges. Each instance raised doubts, as per Callaghan. Each time, the investment represented a novel method for humans to engage with technology that seemed more organic than prior options.

The latest embodiment of this theory is Sandbar, a hardware gadget that Callaghan refers to as a “thought companion” — or, in simpler terms, a voice-activated ring worn on the index finger. Its sole function: to capture and organize your thoughts via voice notes. It’s not designed to rival Humane AI Pin or Oura’s health tracking. “It excels at one task,” Callaghan states. “That single task addresses a fundamental human behavioral requirement that technology currently lacks.”

The intention is not merely to passively capture ambient sound but to be readily available when an idea arises, acting as a kind of cognitive partner. It connects to an app, employs AI, and, as Callaghan asserts, signifies a distinctly different ideology regarding our interaction with intelligence.

What attracted True to the Sandbar creators Mina Fahmi and Kirak Hong was not solely the product. “Upon meeting Mina, we discovered we were completely aligned on vision,” Callaghan remembers. True’s team had spent years contemplating alternative interfaces and making strategic investments in that direction. Consequently, they’d engaged with numerous founders. However, the methodology of Fahmi and Hong — who had previously collaborated on neural interfaces at CTRL-Labs, a startup acquired by Meta in 2019 — was remarkable. “It focuses on what [the ring] facilitates. It’s about the behavior it fosters that we will soon realize we can no longer live without.”

This sentiment echoes Callaghan’s well-known remark about Peloton: “It’s not about the bike.” For some, the bike — even in its initial form — was enticing. Yet Peloton fundamentally revolved around the behavior it encouraged and the community it fostered; the bike merely served as a tool.

This perspective of investing in new behaviors — rather than just new gadgets — also clarifies how True has maintained discipline with its capital. Even while AI startups accumulate hundreds of millions at billion-dollar valuations right from the start, True insists on sticking to its strengths, which involve writing seed investments of $3 million to $6 million for 15% to 20% stakes in startups they often encounter first.

Callaghan mentions that True will gather more funds to support what proves successful, but he has no inclination to pursue billions of dollars. “Like, why? You don’t need that to create something extraordinary today.”

That same cautious mindset shapes his perspective on the overarching AI surge. While he acknowledges (when prompted) that he thinks OpenAI could soon achieve a trillion-dollar valuation, and while he describes this as the most potent surge in computing we’ve witnessed, Callaghan observes cautionary signals in the self-reinforcing financing arrangements backing hyperscalers and their anticipated $5 trillion CapEx expenditure on data centers and chips. “We’re in an incredibly capital-intensive phase of the cycle, which is concerning,” he remarks.

Nevertheless, he remains hopeful regarding where the true prospects lie. Callaghan contends that the most significant value generation is still to come — not in the infrastructure layer but within the application layer, where innovative interfaces will facilitate entirely new behaviors.

Ultimately, it returns to his fundamental investing philosophy, which sounds almost poetic — the kind of astute VC insight that might seem insincere from others: “It should feel frightening and isolating and you should be labeled as insane,” Callaghan reflects on effectively executed early-stage investing. “And it should feel really unclear and ambiguous, but you should be part of a team that you genuinely believe in.” Five to ten years later, he asserts, you will realize if you were onto something.

Regardless, based on True’s history of investing in hardware that numerous others overlooked — fitness trackers, connected bikes, smart doorbells, and now thought-capturing rings — it warrants attention when Callaghan predicts the phone will soon become obsolete. Being proactive is crucial — and the trend trajectories validate his claims: the smartphone market is nearly saturated, with growth barely at 2% per year, whereas wearables — smartwatches, rings, and voice-activated devices — are experiencing growth in the double digits.

A transformation is unfolding in how we desire to engage with technology, and True is strategically placing its investments accordingly.

Displayed above is Sandbar’s Stream ring. For further insights from our discussion with Callaghan, don’t miss the upcoming episode of the StrictlyVC Download podcast next week; new episodes are released every Tuesday.

For a long time, universities and research institutions have served as Europe’s repository for deep tech innovations. Currently, academic spinouts have formed a robust startup pipeline valued at $398 billion — and venture capital is trailing closely behind.

As per Dealroom’s European Spinout Report 2025, 76 of these deep tech and life sciences companies have achieved valuations of $1 billion, $100 million in revenue, or both. Notable unicorns such as Iceye, IQM, Isar Aerospace, Synthesia, and Tekever are now encouraging more investors to support university spinouts.

Just this month, two fresh funds have surfaced that will channel additional investments into talent emerging from European tech universities, expanding a pipeline currently led by Cambridge, Oxford, and ETH Zurich.

PSV Hafnium, based in Denmark, recently finalized its initial fund at an oversubscribed €60 million (about $71 million), concentrating on Nordic deep tech. With branches in Berlin and London, as well as Aachen, U2V (University2Ventures) is aiming for a similar sum for its inaugural fund, for which it has just achieved the first closing.

These two entrants augment the increasing number of European venture firms that consider university spinouts integral to their investment strategies. This category, pioneered by entities like Cambridge Innovation Capital and Oxford Science Enterprises, has now fully matured and diversified.

While it primarily consists of funds associated with one or more universities and institutions, it has also expanded to include independent firms that view spinouts as promising sources of returns — a perception that is justified. Oxford Ionics, acquired by U.S.-based IonQ, was among six spinouts from Switzerland, the U.K., and Germany that generated exits exceeding $1 billion for their investors in 2025.

These exits coincide with rising funding levels. As reported by Dealroom, European university spinouts in deep tech and life sciences are projected to secure an almost record-high $9.1 billion in 2025. This contrasts sharply with the overall VC funding in Europe, which has dropped nearly 50% from its 2021 zenith.

Substantial rounds finalized in 2025 also highlight interest in spinouts across diverse sectors such as nuclear energy — Proxima Fusion — and dual-use drones — Quantum Systems, which is now valued at over $3 billion. Often, these startups draw from specialized research labs, accounting for the notable variety of European locales capable of producing spinouts.

Fostering connections with hubs beyond Oxbridge and leading nations may provide newcomers with a unique edge in sourcing deals. “The Nordic’s research institutions hold extraordinary, untapped potential,” emphasized PSV Hafnium’s partners in a press release.

PSV Hafnium itself originated as a spinout from the Technical University of Denmark (DTU), but is actively investing early-stage funds in other Nordic nations. One of its nine investments thus far was directed towards SisuSemi, a Finnish startup utilizing a decade of research at the University of Turku to develop innovative surface cleaning technology for the semiconductor industry.

It’s positive news for companies like SisuSemi that increased funding is accessible to them. This additional support is complemented by grants, commercialization aid, and enhanced deal conditions, fostering an encouraging atmosphere for Europe’s spinouts. Nevertheless, a persistent challenge remains: growth capital.

As highlighted by the report’s authors, this deficiency “is not an isolated phenomenon concerning spinouts, but a challenge affecting the entire startup ecosystem in Europe.” Still, it is noteworthy that nearly 50% of late-stage financing for European deep tech and life sciences spinouts originates from outside Europe, predominantly from the U.S.

While this proportion has lessened over time, Europe will not fully capitalize on its investments in talent and research unless this dynamic shifts more significantly — a broader problem that requires resolution.

This was meant to be the year that climate technology faded away.

Former President Donald Trump and the Republican Party have actively worked to dismantle key industrial and climate policies under the Biden administration. The European Union has also started to soften its most ambitious objectives.

However, as this year wraps up, the data offers a contrasting perspective on investments in climate and clean energy across the U.S. and Europe. Rather than collapsing, investment in this sector has remained fairly stable compared to 2024, according to CTVC, which is far from the downturn that some had anticipated.

This persistence can partly be attributed to the ongoing challenges posed by climate change. A more significant factor may be that many climate technologies have either become more affordable or superior to their fossil fuel counterparts—or are very close to doing so.

Remarkable cost declines in solar, wind, and batteries continue to propel climate tech forward. While not every emerging technology will experience the same trajectory, these developments suggest that fossil fuels are not infallible, and there are considerable opportunities to invest in companies offering cleaner, cost-effective alternatives.

Data centers continue to be at the forefront

Last year, I forecasted that 2025 would mark the year when climate tech embraced AI and its substantial appetite for energy, a prediction that has largely held true. It’s not entirely unexpected — for the climate tech sector, affordable, clean energy is essential.

Interest in data centers has surged over the last year. Investors surveyed by TechCrunch widely agree that data centers will continue to be a central topic in 2026.

“They are establishing their own financial ecosystem, and the current AI initiatives are gaining enough traction that I don’t foresee the hyperscalers pulling back in 2026,” stated Tom Chi, founding partner at At One Ventures, to TechCrunch.

“I’m consistently hearing about a growing concentration of focus on data centers nearly every day in discussions, particularly with corporate clients,” remarked Po Bronson, managing director at SOSV’s IndieBio, to TechCrunch.

In 2025, the primary concern for data centers was acquiring new energy sources. However, Lisa Coca, partner at Toyota Ventures, believes their focus will shift in 2026. “The conversation around energy for data centers in 2026 is likely to transition from demand to resilience and the necessity to expedite plans to become independent from the grid,” she stated. Independence could help address certain issues faced by data centers, especially resistance from grid operators and the public who are increasingly anxious that new energy demands are driving up their electricity costs.

Nevertheless, the need for additional power persists, and investors noted that geothermal, nuclear, solar, and batteries have all benefited from the surge in interest. “Zero-carbon generation is already among the most affordable energy sources, and the increasing demand for both grid-scale and distributed batteries is speeding up cost reductions more rapidly than anticipated,” mentioned Daniel Goldman, managing partner at Clean Energy Ventures.

Investors acknowledged the possibility of an AI bubble bursting; some expressed doubts about whether such an event would adversely affect the energy sector.

“Could a bubble burst in 2026? Absolutely,” stated Kyle Teamey, managing partner at RA Capital Planetary Health. However, he noted that it is unlikely to disrupt infrastructure plans. “The budget for 2026 has already been allocated. The momentum is in motion.”

Andrew Beebe, managing director at Obvious Ventures, suggested that while the data center bubble may burst around 2026 or early 2027, he does not believe any such bubble exists in electricity generation. “We still require a substantial amount of power, and we’ll utilize it — no build-out bubble in that sector… yet.”

Beyond AI and data centers, Anil Achyuta, partner at Energy Impact Partners, noted that reindustrialization will come into sharper focus this year. “We need to reconstruct supply chains for systems requiring multiple components and complex processing flows,” he said, highlighting robotics, batteries, and power electronics as key areas.

The ongoing pursuit of energy

Thanks to a continuous stream of new data center announcements, energy-focused startups have received a boost this year, particularly those involved in nuclear fission. In recent weeks, nuclear startups have reported funding rounds exceeding $1 billion, raising speculation that many will pursue SPAC or traditional IPO processes in 2026.

“Nuclear is very much in vogue at present,” remarked Teamey.

However, it will take time for nuclear energy to make a significant impact on electricity demand. In the interim, technology firms and data center developers have been turning to solar and batteries as cost-effective, rapidly implementable energy sources. Grid-scale batteries, in particular, have seen record deployments in 2025. With the emergence of alternative battery chemistries such as sodium-ion and zinc, costs are expected to fall further, promoting widespread adoption.

“We expect to see growth in 2026 with new innovations in [battery] chemistry and business models,” stated Leo Banchik, director at Voyager. “A significant lesson from past mishaps was the need to develop gigafactories after validating demand or achieving superior unit economics compared to existing models. The current wave is more disciplined.”

Several investors are optimistic about geothermal stepping up to fill impending gaps in the coming years. They view enhanced geothermal as a relatively advanced technology poised for large-scale deployment in 2026.

“Geothermal will closely follow solar in terms of new generation,” stated Joshua Posamentier, managing partner at Congruent Ventures. “Natural gas assets are expanding at a steady pace. There isn’t much new capacity in turbine manufacturing coming online, and they are fully utilizing what they can. Geothermal is poised for exponential growth.”

While AI is driving demand, innovations and businesses that venture beyond data centers are expected to gain the most, according to Laurie Menoud, founding partner at At One Ventures. “Data centers are one component of the demand, not the entirety of the market.”

Which startup is poised for a public offering in 2026?

Not everyone concurred or was willing to offer a prediction. However, among those who did, many highlighted nuclear or geothermal startups as likely candidates for going public, whether via IPO or SPAC.

The startup most frequently mentioned was Fervo, the enhanced geothermal company that recently secured a $462 million funding round. The firm is widely considered a frontrunner in the industry and is currently constructing a 500-megawatt facility in Utah that is expected to serve as a blueprint for forthcoming power plants. Accessing the public markets would enable the company to bolster its resources for additional projects.

Trends to monitor

Apart from data centers, investors are keen on a variety of technologies and sectors, including essential minerals, robotics, and software designed to manage the electricity grid.

“We need to focus more on grid execution as a vital category,” said Amy Duffuor, general partner at Azolla Ventures. “The hidden successes are companies that produce software, hardware, and supply-chain solutions to expedite interconnection, planning, and deployment, helping utilities progress on projects.”

Resilience and adaptation will be key themes in 2026, according to Coca of Toyota Ventures and Posamentier of Congruent Ventures. Achyuta at EIP identified a potential application: systems that enable robots to bury electrical transmission lines more swiftly and cost-effectively than humans, thereby reducing wildfire risks and enhancing the grid’s reliability.

Beebe, at Obvious Ventures, noted that electric trucking will also be a space to watch. “One of the major developments of 2026 will be Tesla Semi’s releases and specifications. The range and pricing of that vehicle will transform the industry in ways comparable to the Model S or 3.”

AI is likely to play a significant role in transforming climate tech. “We anticipate substantial innovation when AI interacts with the physical world in 2026, affecting both infrastructure and consumer applications,” stated Matt Rogers, founder at Incite and Mill. “The integration of AI with smart hardware and physical infrastructure will catalyze the transformation of trillion-dollar industries from manufacturing to life sciences to food systems.”

Moreover, keeping an eye on technologies that have previously been deemed impractical could be beneficial, remarked Bronson at SOSV. “When investors eventually grow weary of a sector and conclude that it is unlikely to yield returns, that’s precisely when true breakthroughs occur,” he commented.

Explore further

Below are the detailed remarks from the investors who participated in TechCrunch’s survey, listed alphabetically. Click the link to jump to a specific response.

• Anil Achyuta, partner at Energy Impact Partner
• Leo Banchik, director at Voyager
• Andrew Beebe, managing director at Obvious Ventures
• Po Bronson, managing director at SOSV’s IndieBio
• Tom Chi, founding partner at At One Ventures
• Lisa Coca, partner at Toyota Ventures
• Amy Duffuor, general partner at Azolla Ventures
• Daniel Goldman, managing partner at Clean Energy Ventures
• Laurie Menoud, founding partner at At One Ventures
• Joshua Posamentier, managing partner at Congruent Ventures
• Matt Rogers, founder at Incite and Mill
• Kyle Teamey, managing partner at RA Capital Planetary Health

Anil Achyuta, partner at Energy Impact Partner

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

Reindustrialization beyond data centers will emerge as a significant theme. We need to restore supply chains for systems that entail multiple components and intricate flowsheets. For instance, advancing next-generation robotics to tackle labor shortages and national security challenges will necessitate integrated supply chains. Technologies like batteries, power electronics, fuel cells, gas turbines, and even home manufacturing are examples of markets/technologies that will need reinvention within the value chain.

Another area to watch is AI-powered physical science. While companies like Zanskar (predictive AI for geothermal) and Fabric8Labs (generative cooling for data centers) have demonstrated promise, there haven’t been many visible breakthroughs yet. Nonetheless, the talent pool addressing these challenges is impressive and could lead to exciting advances.

Where is the largest opportunity to find or place power on the grid?

Gas turbines deliver reliable capacity and remain a choice for many major players deploying data centers. Beyond that, batteries—especially sodium-ion—represent one of the most economical, near-term solutions at the grid scale. I am optimistic about the progress in this technology, and pairing solar with batteries (as firms like Peak Energy are doing) remains an attractive strategy. Next-gen geothermal also shows significant promise, but the timelines are akin to those of nuclear energy—powerful yet taking about a decade to reach full capacity.

Additionally, algorithmic solutions that unlock new power using existing infrastructure (e.g., Gridcare, ThinkLabs AI) and optimize workloads (e.g., Emerald AI) can enhance grid efficiency. Innovations such as applying optical coatings to transmission lines to minimize losses, are being developed by AssetCool (a company within the EIP portfolio).

Which climate tech or clean energy startup is most likely to IPO in 2026?

Fervo Energy, Commonwealth Fusion, and Redwood Materials would be my personal guesses, though I may be mistaken.

Which technologies do you believe will be ready for larger-scale deployment in 2026?

Sodium-ion batteries for grid-scale storage are already in deployment and will see rapid acceleration in 2026. Another technology to keep an eye on is solid-state transformers (note that Heron Power is a part of the EIP portfolio). The industry is progressing more swiftly than expected and is scaling similarly to semiconductors, though full-scale production may require more time.

What trend or technology should we focus on more?

One emerging trend is the underground construction of transmission lines. Advances in robotics could facilitate a quick, cost-effective approach that greatly diminishes wildfire risks, thus mitigating the substantial carbon emissions connected to such incidents.

Distributed power, warmth, and computation are the final category of trends we are tracking with interest for 2026.

Leo Banchik, director at Voyager

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

Data centers will persist in driving record energy consumption as AI workloads expand. Despite concerns about overbuilding, it’s unlikely there will be significant stranded capacity — as computing power becomes less expensive and more accessible, we will continue to uncover new applications for it. An intriguing shift is seen in hyperscalers who are starting to distinguish between clean energy sources – robust versus intermittent, location, and scalability – instead of merely considering megawatt-hours. This is already emerging in customized offtake agreements and on-site supply strategies.

Fission and geothermal are expected to maintain momentum fueled by both private investments and federal support. As geopolitical tensions rise, fusion may also gain traction for increased federal backing, although significant deployment on a high-capacity grid level is still years away.

Alternatives to natural gas peakers will likely gain momentum too—employing newer turbines and modular designs, coupled with integrated carbon capture, as grids manage increased peak demands from AI.

Where is the greatest opportunity to find or place power on the grid?

The expansion of solar and batteries will continue, given their strong economics. For persistent, dispatchable baseload power, we anticipate expansion in fission, geothermal, and peaker alternatives, like modular gas turbines with carbon capture. Additionally, there’s a grid-edge opportunity worth monitoring: significant facilities acquiring dedicated baseload on-site rather than adding to grid congestion.

Which climate tech or clean energy startup is most likely to IPO in 2026?

Most likely in fission or geothermal. These firms have secured substantial capital and negotiated robust offtake agreements with hyperscalers and utilities. With multibillion-dollar project pipelines requiring ongoing growth financing, several could venture into public markets in 2026.

Which technologies do you think will be prepared for larger scale deployments in 2026?

The deployment of energy storage is speeding up across residential, commercial, industrial (including backup for data centers), and grid-scale applications. Domestic supply chains, encompassing second-life battery systems, are gaining traction for stationary storage. Anticipate growth in 2026 with new approaches regarding chemistry and business models. A critical lesson learned from previous setbacks was the urgency of scaling gigafactories post demand proof or when achieving better unit economics than conventional models. The emerging wave is more disciplined.

Industrial heat pumps and thermal storage solutions for steam and process heat are becoming less expensive to operate than gas boilers in numerous regions and applications, especially where waste heat is available and electricity rates are favorable.

We’ll also see substantial expansion in critical minerals and battery materials initiatives — lithium, rare earths, magnesium refinement; battery component and cell manufacturing; copper recycling — emerging with federal support as supply chain security turns strategic.

What trend or technology should we pay more attention to?

Software and AI enhancing physical infrastructure: real-time factory intelligence to improve energy efficiency and production yields, AI-based design tools that expedite product development cycles, grid management software coordinating intermittent renewables with storage and dispatchable power.

Companies adopting a clean-slate strategy to reimagine foundational technologies — think a SpaceX-style reevaluation of components once thought to be resolved issues. Developing motor designs that eliminate reliance on rare earths, modern manufacturing techniques for grid infrastructure like transformers, advanced materials processing that significantly reduces costs while enhancing quality. Advances in robotics further support these cost trajectories, making U.S. manufacturing competitive again.

Finally, dual-use climate technologies with superior unit economics that inadvertently strengthen domestic supply chains. Defense and industrial policy continue to back these not merely for climate-related reasons but because they present cost advantages and ensure supply security.

Andrew Beebe, managing director at Obvious Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

Data centers will once again take precedence. There will, however, be heightened discussions about a potential build-out bubble (focused on data centers, not electricity generation). We’ll grapple with the dual reality of excessive funds/debt allocated to data centers, and the speculation bubble will likely burst (possibly early 2027). Concurrently, we will still necessitate substantial power, which will be utilized — no build-out bubble in that sector… yet.

Where is the largest opportunity to locate or place power on the grid?

In terms of power generation: geothermal in the near term. Fission in the mid-term. Fusion in the long-term, over ten years. Regarding siting: The aforementioned technologies can be deployed anywhere, but primarily in western states for geothermal electricity. For batteries—PJM [the grid covering the mid-Atlantic west to parts of Illinois] and Texas.

Which climate tech or clean energy startup is likely to IPO in 2026?

Fervo is among the promising candidates from venture backing.

Which technologies do you believe will be ready for larger-scale deployment in 2026?

Geothermal and grid-scale batteries.

What trend or technology should we focus on more?

Grid software and electric vehicle trucking. A key highlight of 2026 will likely be Tesla Semi’s specifications and release. The range and pricing of this vehicle will significantly reshape the industry, reminiscent of the impacts made by the Model S or 3.

Po Bronson, managing director at SOSV’s IndieBio

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

I still observe an ever-growing concentration of effort and emphasis on data centers nearly every single day during meetings, particularly with corporates. This is somewhat driven by “picks and shovels” companies who are strategizing how to become more significant players/integrated instead of merely being a purchased component.

A related term I encounter more often is power density and/or specific power (power-to-weight ratio), as many corporates anticipate or plan a transition of their energy divisions into robotics. Duncan Turner is our resident expert on this topic.

Which climate tech or clean energy startup is most likely to IPO in 2026?

I don’t have a climate tech company in my portfolio lined up for public offerings in 2026. Tidal Vision has targeted 2027. That’s my closest prediction. I prefer not to speculate on other VC portfolios, even though I have my opinions, as I shouldn’t express views where I’m only partially informed.

Which technologies do you think will be ready for larger-scale deployment in 2026?

For 2026, my fastest-scaling companies include Tidal Vision and Voyage Foods, which has taken over a General Motors facility in Ohio.

What trend or technology should we pay more attention to?

Regarding what to monitor more closely, I’ll repeat Duncan’s accomplishments — his initiatives with the Plasma Forge are, in my opinion, going to be highly compelling and will ensure rigorous study in the field.

Additionally, I consistently feel that real breakthroughs often occur when investors grow weary of a sector and conclude that it’s unlikely to thrive. I recall this lesson from 1999 when there was speculation about whether the search market would be dominated by Yahoo, AltaVista, Excite, Lycos, or Infoseek.

I sense this trend in my personal investments. I recently conveyed this sentiment to AgFunder; however, the VC landscape seems to exhibit a bias that anticipates multiple winners in sectors projected for growth. In reality, most markets don’t accommodate multiple winners, and a single company emerges victorious.

Tom Chi, founding partner at At One Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

Expect significantly more discussion surrounding data centers in 2026. They are crafting their own financial ecosystem, and the current momentum in AI initiatives suggests that I do not foresee hyperscalers retracting in 2026.

Where is the biggest opportunity to find or place power on the grid?

Budgets for hyperscalers range from $50 billion to $100 billion, covering power, chips, and more. The costs associated with chips are substantial enough that stakeholders are willing to allocate additional funds to secure power on the grid sooner, as the losses from chip depreciation exceed most incremental additions to their power expansion budgets.

Which climate tech or clean energy startup is most likely to IPO in 2026?

The IPO market remains a bit vague, and most individuals don’t reveal the timing of their public offerings.

Which technologies do you think will be prepared for larger-scale deployment in 2026?

Companies like Fervo are at an interesting pivotal moment. One of our portfolio companies, Provectus Algae, is also at a crucial juncture.

What trend or technology should we prioritize more?

We’ve observed a significant swing away from heavily capital-intensive endeavors in industrial decarbonization not tied to AI. It is essential for our collective future, notwithstanding its temporary period of diminished visibility.

Lisa Coca, partner at Toyota Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

From Toyota Ventures’ perspective, we believe that the energy dialogue around data centers in 2026 will transition from demand to resilience, emphasizing the urgency to advance plans for grid independence.

Where is the biggest opportunity to find or place power on the grid?

We are convinced that the most promising investment opportunities lie in firm, dispatchable, and scalable carbon-free energy solutions. We have actively committed to technologies enhancing baseload power, such as geothermal and nuclear, through our portfolio companies like Rodatherm and Natura Resources. For essential grid flexibility, we are investing in advanced, long-duration energy storage battery technologies with our backing of e-Zinc.

Which climate tech or clean energy startup is most likely to go public in 2026?

We foresee nuclear power leading in terms of IPOs and SPACs during 2026.

Which technologies do you think will be ready to deploy at larger scales in 2026?

This is a challenging question since the outcome may hinge on the evolution of the capital stack. A substantial number of climate tech companies across various sectors are on the brink of larger-scale deployment. The key challenge lies in securing financing for first-of-a-kind (FOAK) projects to mitigate risks in the crucial step of transitioning from initial prototypes to scalable implementations.

What trend or technology should we focus on more?

Our team anticipates that resilience and adaptation will continue to be prominent in 2026. The Toyota Ventures portfolio exemplifies this: BurnBot focuses on wildfire mitigation, ZymoChem enhances supply chain resilience through sustainable materials, and Alora creates adaptable resource solutions.

Amy Duffuor, general partner at Azolla Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

I predict that the energy discourse will shift from power generation to the speed of power delivery. Interconnection timelines, permitting, and physical grid constraints remain bottlenecks, leading data centers to increasingly rely on hybrid strategies combining grid power, storage, and demand flexibility to meet timelines.

Where is the biggest opportunity to find or place power on the grid?

One opportunity lies in grid-ready locations, such as sites with existing transmission and substations. Anything that shortens interconnection timelines currently holds immense value due to limited access to the grid. I am also interested in wireless power transmission, despite its nascent stage.

Which climate tech or clean energy startup is most likely to IPO in 2026?

Fervo Energy has generated considerable talk recently…!

Which technologies do you think will be ready for larger-scale deployment in 2026?

Long-duration energy storage technology companies are set to progress from initial pilots to demonstrations and subsequently to repeatable deployments. We’re particularly enthusiastic about our portfolio company Noon Energy.

What trend or technology should we emphasize more?

Grid execution should gain more prominence as a category. The unnoticed victors are firms that produce software, hardware, and supply-chain solutions that expedite interconnection, planning, and deployment, empowering utilities to effectively advance projects.

Daniel Goldman, managing partner at Clean Energy Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

We anticipate an uptick in deal-making within the data center/hyperscaler sector that encompasses the following aspects:

1. Organized power off-take agreements that combine behind-the-meter and utility-related infrastructure to enhance pricing and reliability;
2. More actions at federal, ISO/RTO, and state levels to expedite the deployment of energy assets while balancing tariff structures to avoid imposing increased costs on voters;
3. Mergers and Acquisitions focused on technology optimization, inclusive of resources such as geothermal, nuclear, critical minerals, and downstream hardware and software products that facilitate the digitalization, decarbonization, and distribution of energy resources and load management, a focal area for our firm and venture capital at large.

While we do not predict an overall “bust cycle” for data center and hyperscaler development activities, we foresee some rationalization in development and the implementation of efficiency options to reduce capacity needs.

Where is the largest opportunity to find or place power on the grid?

The most significant immediate opportunity — and challenge — lies in enhancing the grid itself. Modernizing the grid through digital platforms, decarbonization, and decentralization will unlock cost efficiencies, optimize existing infrastructure, and better integrate large-scale distributed energy resources — likely not groundbreaking information here. Zero-carbon generation is already among the most economical energy sources, and the rising demand for both grid-scale and distributed batteries is catalyzing cost reductions more rapidly than expected.

We foresee this trend persisting despite recent policy shifts evident in the IRA, as numerous venture capital-supported companies have the potential to make a significant impact on the grid as they expand and gain market adoption. The disruption is underway!

Which climate tech or clean energy startup is most likely to IPO in 2026?

Factorial appears to be a leading candidate following its plans to de-SPAC in 2026. Its trajectory indicates that companies with solid customer traction in expansive markets, clear cost and performance benefits, and fast-growing revenues are well-positioned for public offerings. (This SPAC market contrasts sharply with that of 2020, where entrants such as QuantumScape lacked significant revenues and may not have been adequately prepared for public markets.)

Beware of an increase in companies rushing to public markets on hype rather than solid fundamentals.

Beyond Factorial, several companies in energy storage, generation, and critical mineral sectors are nearing scalable revenue heights to facilitate access to low-cost public market capital (a genuine advantage); however, in the minerals sector, we may see a surge in mergers and acquisitions preceding any public offerings in 2026.

Which technologies do you think will be ready to deploy at larger scales in 2026?

Projects in energy storage, sustainable aviation fuel (SAF), critical minerals, and manufacturing facilities throughout the energy supply chain will attract significant investments in the U.S., supported by manufacturing tax incentives and attractive market prospects, despite potential federal policy challenges.

In 2025, we thoroughly examined and developed new risk transfer solutions for FOAK project developers. We foresee commercial lenders and private credit beginning to engage significantly in this domain with support from insurance underwriting and catalytic capital. There’s an expectation of a rise in projects and debt financing to back early-stage commercialization, which historically has only trickled in. This is a crucial enabler of the broader scale necessary across the industry.

What trend or technology should we keep a close eye on?

For western markets to be competitive against China’s manufacturing and innovation capabilities, financial innovation is CRUCIAL. Global markets must deploy an estimated $3 trillion to $9 trillion annually through 2050 on climate-related technologies and project implementations if we aim to mitigate global temperature increases and compete internationally in these arenas. Climate-related investment globally reached only $2 trillion in 2024 and is on a trajectory to match that figure in 2025.

To boost the deployment rate, we must persuade public and private investors that the risk-return profile is favorable enough to deploy capital across the climate capital stack — this includes early-stage ventures, growth-stage ventures, private equity, commercial lending, and private credit, together with infrastructure. Presently, our sector is not drawing enough capital; simply put, risk levels need to diminish or returns need to increase.

There are opportunities for employing risk-sharing strategies to optimize capital structures and lower the financing costs for new technology initiatives, which will also accelerate their descent along the cost curve. Viable solutions include technology and performance risk insurance, surety bonds to handle construction risks, pooling offtake agreements among buyer groups (e.g., hyperscalers for clean power or airlines for SAF), knobby financing gaps in construction, and beyond. Clean Energy Ventures has spent time in 2025 identifying new risk transfer solutions, collaborating closely with our colleagues in finance and insurance sectors. We believe 2026 will witness more innovative financing frameworks facilitating the swifter scaling of climate technologies.

Attention should also shift to cost trends. The impact of AI is emerging but not gaining widespread recognition within climate technology. Within larger corporations and smaller startups, AI is driving cost reductions, prompting quicker innovation in complex facilities and supply chains. This trend is visible across chemicals, mining and refinement, power generation and grid optimization, and manufacturing (steel, cement), recycling, and waste management, among others.

We are still in the early phases of witnessing AI’s effect on cost curves across diverse commodities and sectors. As we discuss the upward influence of power prices driven by AI infrastructure needs, it’s essential to remember that AI will concurrently remodel industries worldwide and lower production costs.

Laurie Menoud, At One Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

There’s undeniably been substantial enthusiasm around everything associated with data centers, energy production, storage, transmission, and cooling. However, from a venture capital standpoint, constructing and scaling data centers does not align with startup timelines. It’s a prolonged struggle involving permits, substations, and grid enhancements. To quantify it, hyperscale data centers currently require about three to six years to obtain permits and connect in the U.S. In some contexts, interconnection alone can extend beyond five to seven years. Therefore, in 2026, I expect to observe sustained progress for energy companies not only linked to data centers but capable of expanding into commercial and industrial applications and front-of-the-meter initiatives. Data centers represent one demand influencer, not the entirety of the market.

Additionally related is the supply of essential metals, which I remain highly focused on: mining, extraction, refinement, and recycling. This is critical not only for data centers (particularly copper) but also for EV batteries, which rely on materials such as lithium, nickel, manganese, and cobalt.

Where is the most significant opportunity to discover or place power on the grid?

In regions where thermal (coal and gas) and nuclear facilities are retiring since these locations already possess strong grid connections. Rapidly deploying new clean energy generation at these sites could significantly hasten project timelines. In the U.S. alone, over 60 GW of coal capacity has been retired since 2015, with an additional 40+ GW scheduled for retirement by 2030. Each of these retirements frees up a transmission node that took decades to establish. Harnessing or reusing existing interconnection can often be the deciding factor in establishing whether a project timeline is two years versus eight years. This would facilitate the installation of next-gen nuclear facilities like Stellaria, which reduces long-lived waste, lower capital expense, operational costs, minimizes deployment time, and extends fuel usage, or geothermal solutions like Factor2 Energy, utilizing underground CO2 reservoirs to lessen deployment location constraints.

The same principle applies to industrial sites (chemicals, steel, refineries) that have existing oversized grid connections. These sites are endeavoring to increase production and add storage, achieving a significant challenge: interconnection. If rapid electrification of heavy industries is the goal, focus should be on areas where the grid already exists.

Which climate tech or clean energy startup is most likely to IPO in 2026?

While none can be definitively known, I would closely watch the battery recycling and circular supply chain for critical materials sector. The prices of lithium, nickel, and cobalt are incredibly sensitive to geopolitical forces, and recycling offers a lower-risk, domestic supply for the U.S.

Which technologies do you believe will be ready for larger-scale deployment in 2026?

Relectrify is gearing up to deploy their battery systems at commercial production scales in 2026, targeting a cumulative capacity of 100 MWh. Their strategy uses semiconductor circuitry at the cell-level to control battery cells individually at high frequency, generating a direct alternating current waveform without requiring an inverter anymore, which presents clear wins in CAPEX, enhanced battery lifespan, and lower OPEX through precise identification and replacement of malfunctioning cells. This process is already underway.

Additionally, grid-scale energy storage beyond lithium is essential, driven by both AI data centers and renewable advancements. Without significant storage, maintaining 24/7 clean energy beyond nuclear and hydro is unfeasible. Globally, stationary storage is projected to escalate from approximately 45 GWh in 2023 to hundreds of GWs by 2030. Some of the cutting-edge technologies are already prepared today.

Battery recycling and circular supply chains with automotive manufacturers (recycled lithium, nickel, cobalt, and copper returning for new batteries) are progressing and liquidating into larger scales as well. The year 2026 will undoubtedly emphasize acceleration. Ascend Elements has already constructed North America’s largest lithium-ion battery recycling facility, achieving the first production of recycled lithium carbonate. Presently, most of the U.S. supply of lithium carbonate is imported, primarily from Argentina and Chile. Securing metal supply with lower-cost recycled content represents a considerable advantage.

What trend or technology should we pay additional attention to?

Firms such as Chemfinity, which have the potential to position domestic metal refining on par with China, are noteworthy, as is anything connected to mining, extraction, refinement, and recycling of essential metals for data centers and electric vehicles. Copper is THE metal for data centers, utilized in power cables, busbars, transformers, and cooling loops. A single gigawatt of data center capacity requires tens of thousands of tonnes of copper. Approximately 40% to 45% of the world’s copper refining occurs in China, followed by Chile, mirroring the geographic distribution seen in lithium refining and precursor battery chemistry. Conversations around energy security should focus on this reality.

Joshua Posamentier, managing partner at Congruent Ventures

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

The push for growth will persist, but the focus will shift from dramatic gigawatt announcements to construction, commissioning, and grappling with the harsh realities of interconnect delays and permitting challenges. Consequently, there will be a significant emergence of “bring your own generation” and “demand flexibility.”

The era of Application Specific Standard Products (ASSPs)/Application Specific Integrated Circuits (ASICs) for AI data centers will commence earnestly. Although GPU demand will continue to grow, this pace is expected to slow in favor of more specialized chips that exhibit far better efficiency, particularly for inference tasks. This shift will progressively decouple data center electricity consumption from token generation.

The unit economics of the leading foundational AI firms will leak, revealing unfavorable indicators. Nevertheless, these sectors will likely convince investors to support them until they achieve positive unit economics, a breakthrough that will materialize sooner than anticipated. However, the allure will wane; these firms will be classified differently from the high-margin SaaS companies from earlier market cycles.

Data centers will enhance their roles as grid participants through adaptability, load management, and power quality, thanks to technology adoption leading to much faster interconnect times compared to other large energy loads — the more technology they embrace, the quicker they will be connected.

In 2026, we will likely witness the first off-grid, world-scale data center notice to proceed (NTP) [a letter instructing a contractor to commence work]. At the same time, expect stronger NIMBY (Not In My Back Yard) resistance to nearby data centers due to a myriad of concerns ranging from energy expenses to size and water usage.

Where is the biggest opportunity to find or place power on the grid?

Nuclear fusion! I anticipate that we will observe the first net gain analysis (i.e., deuterium fusion where it would be Q>1 if it were tritium) within a startup reactor.

Geothermal will be closely following solar in terms of new generation to increase gas asset deployments.

Which climate tech or clean energy startup is most likely to IPO in 2026?

Whether through SPACs or IPOs? Numerous opportunities lie ahead.

What technologies do you think will be ready to deploy at larger scales in 2026?

Geothermal technologies for electricity and district heating. The friction associated with scaling single-site geothermal heat loops still remains too significant to see rapid enhancements beyond current levels. However, thermal energy storage for load management in industrial environments will gain attraction.

What trend or technology should we be paying more attention to?

Robotics (excluding humanoid types) are taking over numerous labor-intensive sectors; they will significantly influence various industries such as manufacturing, agriculture, waste management, and more in 2026.

Logistics and manufacturing efficiency are also in the spotlight: electrification, efficiency enhancements, onshoring, and AI are exerting pressure on emissions across approximately half of the economy, primarily driven by economic, rather than purely impact, motivations. While providing them enduring clarity on lower costs compared to conventional fuels would entice buyers, it pertains to everything from electrified autonomous trucking to electric autonomous rail and hidden terminals.

The demand for resilience technology is set to accelerate significantly. Insurance costs stemming from climate risk are escalating faster than any other expenditures for homeowners, which will also affect commercial enterprises. Businesses and individuals will actively begin investing in resilience, confronting intensifying climate change and extreme weather events, as well as the aging infrastructure and the transition from a centralized to a decentralized resource paradigm.

Matt Rogers, founder at Incite and Mill

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

Data centers will function as pivotal factories for the next wave of AI innovation, driving America’s societal and economic landscape. I envision that local governments will assume a more active role in 2026, challenging hyperscalers to deliver solutions aligned with community needs, facilitating municipal partnerships that enable quicker construction. Energy affordability is a pressing concern, and reversing the trend of rising costs from 2025 is vital.

Where is the largest opportunity to discover or locate power on the grid?

The initiative for decentralized infrastructure solutions, including rooftop solar, energy storage, and distributed energy resources like heat pumps and smart thermostats, originates from within households in 2026. These options are accessible today and can be swiftly activated in a cost-effective manner, culminating in minimal disruption to grids throughout the nation. The permitting process is quicker than constructing new centralized power plants.

Should sufficient individuals adopt these readily available solutions, America’s grid could better manage the increased capacity from the expansion of data center AI.

Furthermore, there’s a unique opportunity for collaboration: AI innovators and major tech firms seek a quicker, more reliable route for development. Meanwhile, local and state governments are interested in affordability, economic investments, and resilience. This collaborative effort puts communities in a powerful position to negotiate swift permitting and adaptable construction timelines for private-sector-fueled economic revitalization, tax income, and job creation.

By empowering households with efficient tools to reduce utility expenses, hyperscalers will find it easier to access the power required to run AI data centers, resulting in significantly quicker responses. This endeavor is not abstract; rather, it is tangible and practical.

Which technologies do you think will be ready for larger-scale deployment in 2026?

Mill. We plan to implement more Food Recyclers, and at a larger scale than ever before. We are strategizing additional partnerships to persuade more households, businesses, and communities to perceive food as a resource that ought to return to the food system rather than being discarded for weeks in dumpsters or linger for years in landfills.

Robotics will attract considerable funding and private sector interest in 2026. Simultaneously, builders will steer away from hype surrounding humanoid models, moving towards practical robots tailored for specific tasks, thereby improving life quality. The robotics landscape will focus more on functions akin to Roomba than relatable human companions, benefitting from advancements in both affordability and technology risk management.

What trend or technology should we concentrate more on?

In 2026, expect to see significant innovation where AI intersects with the physical world, impacting both infrastructure and consumer application layers. Merging AI with intelligent hardware and physical infrastructure will facilitate the transformation of trillion-dollar sectors, spanning manufacturing to life sciences to food systems. For AI to prop up the future without overwhelming grids or negatively affecting communities, it must be integrated with tangible hardware that addresses challenges and enhances everyday life. The benefits will extend to both AI-driven smartphones and AI-optimized waste facilities.

Kyle Teamey, managing partner at RA Capital Planetary Health

Data centers have dominated conversations about energy in 2025. What should we expect in 2026?

I anticipate that data centers will still hold prominence in 2026. However, my perspective is slightly colored by past experiences from the previous AI cycle—it feels reminiscent of similar conversations. Yet, the scale of investment is remarkably greater. The amount of attention devoted to it is exponentially more significant. Therefore, the resolution of this situation may take considerable time to unfold.

The expenditure allocated for 2026 has already been earmarked. The momentum has begun. Could a bubble burst in 2026? Certainly. Yet, such repercussions may take time to materialize. It could require several months up to a year to truly manifest. It is a formidable task to halt momentum mid-course and attempt to retrieve funds.

Where is the most substantial opportunity to find or place power on the grid?

There has been considerable discourse on this topic, yet the demands related to data and its requirements grow at an exponential rate. In contrast, the power scale increases in a linear fashion. As a result, it may take an extended period for the physical infrastructure to keep pace with the mounting data demands. The possibilities are quite vast — whether it pertains to advancements in power generation, storage, transmission, or distribution; improving grid functionality encompasses an extensive list.

Given the current bull market in electricity, it’s challenging to determine how long this phenomenon has persisted—possibly for about a century? In our assessment, the opportunities abound across various areas. Recent months have seen some entities going public, and it is fair to predict an uptick in this trend.

Which climate tech or clean energy startup is poised for a public offering in 2026?

I foresee a noticeable increase in public companies emerging from the power generation domain in 2026, and this could encompass a wide range of players. Nuclear energy is notably gaining traction currently, and we can likely expect an influx of such companies. Geothermal energy also stands a good chance, and numerous intriguing firms engaged in project development and execution could potentially pursue public status as well. It’s not solely technology firms; rather, those involved throughout the supply chain may also come to the forefront.

Which technologies do you believe will be set for larger-scale deployment in 2026?

Particularly nuclear fission companies have considerable potential for growth. While this may represent a bubble, there is a chance that if certain companies find success in bringing their projects to fruition, it could lead to increased capital inflow to facilitate further rapid expansion.

What trend or technology should we be monitoring closely?

No groundbreaking technologies have surfaced recently that I view as game-changers deserving universal attention. Presently, the emphasis revolves more around scaling many of these technologies. Achieving rapid scale presents a remarkable opportunity.

The various trends are driven not solely by manufacturing demand but also by the regionalization of resources, which is generating substantial requirements for labor, resources, and beyond. The demand is evident across countless sectors.

In various respects, 2025 marked the real emergence of AI dictation applications. While dictation software has existed for several years, its previous iterations were often slow and imprecise — particularly if one did not adhere to specific accents and clear enunciation.

However, developments in large language models (LLMs) and speech-to-text algorithms have significantly enhanced the ability of these systems to interpret spoken language with improved contextual understanding to format the written text. Developers have also integrated functionalities that automatically format text, eliminate filler words, and disregard mistakes, resulting in text that requires minimal revisions.

With the increasing prevalence of AI technologies, numerous such applications have emerged in the marketplace. Consequently, we have compiled a selection of the most effective and beneficial dictation apps of this year.

Wispr Flow

Wispr Flow is a well-supported AI dictation tool that allows users to input custom vocabulary and instructions for dictation. It is available as native applications on macOS, Windows, and iOS, with an Android version currently in development.

The application enables you to tailor the transcription process by selecting from “formal,” “casual,” and “very casual” writing styles for various contexts, including personal messages, professional work, and emails. Furthermore, if you utilize vibe-coding tools like Cursor, a feature can be activated to automatically identify variables or tag files in conversation.

You can dictate up to 2,000 words each month without charge on any desktop version, while the iOS version permits 1,000 words monthly at no cost. Subscription options provide unlimited transcription capabilities, beginning at $15 per month.

Willow

Willow promotes itself as a significant time-saver for individuals who prefer not to type. In addition to standard functionalities like automatic editing and formatting, it features a capability that leverages large language models to produce substantial blocks of text from just a few dictated phrases.

Willow also takes a more privacy-centric approach to AI-enhanced note-taking by storing all transcripts directly on your device, allowing users to opt out of model training if they wish. It also supports custom vocabulary additions to adapt to industry-specific language or local dialects.

Willow allows users to dictate 2,000 words per month for free via its desktop application. Individual subscription plans commence at $15 monthly, providing unlimited dictation and facilitating the app’s capacity to remember your unique writing style.

Monologue

For those prioritizing privacy, Monologue offers the option to download its model for on-device transcriptions, bypassing the need to transmit data to the cloud. Additionally, users can personalize the tone of voice to align with the applications they utilize alongside it.

Monologue allows for dictation of 1,000 words per month at no charge, with subscriptions available for $10 monthly or $100 annually. For frequent users, the company provides a unique Monokey to complement the app.

Keyboards are so 1983.

You only need one key—Monokey, the limited edition device that turns your voice into text in Monologue.

We’re giving away 10, along with a free annual subscription to Monologue. pic.twitter.com/nXuz1ll2LU

— Monologue (@usemonologue) November 18, 2025

Superwhisper

Superwhisper primarily functions as a dictation application but also offers transcription services for audio or video files. The app empowers users to select and download AI models, including its own varying in speed and accuracy, as well as Nvidia’s Parakeet speech-recognition models.

It also allows users to create custom prompts to direct the output. You can conveniently view both processed and unprocessed transcripts integrated with the system keyboard.

[embedded content]

The fundamental voice-to-text functionality is free, and you can explore Pro features such as translation and transcription for 15 minutes. The paid tier grants you access to employ your own AI API keys, plugging in both cloud-based and local models without limitations.

The monthly subscription costs $8.49, the annual plan is priced at $84.99, or you can opt for a lifetime subscription at $249.99.

VoiceTypr

The VoiceTypr application adopts an offline-first, subscription-free model, permitting you to utilize local models for transcription. A GitHub repository is also available for those interested in hosting and managing the open-source version independently. VoiceTypr accommodates over 99 languages and operates on both Mac and Windows platforms.

The app provides a three-day free trial, after which you can purchase a lifetime license. Prices are set at $35 for a single device, $56 for two, and $98 for four devices.

Aqua

Aqua is yet another Y Combinator-funded voice-typing client for Windows and macOS, touted as one of the swiftest tools in its category regarding latency.

In addition to managing grammar and punctuation, Aqua enables users to autofill text by dictating phrases — for example, saying “my address” will prompt Aqua to type in your address.

The application also features its own speech-to-text API for integration with other applications.

[embedded content]

The free version allows for 1,000 words per month, while payment plans commence at $8 per month (with annual billing) and provide unlimited words along with 800 custom dictionary entries.

Handy

Handy is a free, open-source transcription tool compatible with Mac, Windows, and Linux. Though the application is quite basic and lacks extensive customization, it serves as a solid option for those wishing to begin using their voice more frequently without incurring costs.

The application includes a simple settings menu that allows for toggling push-to-talk and modifying the hotkey that activates the transcription function.

Typeless

Typeless presents itself as another app in this sector with a generous free word limit. The company asserts that it does not store any data or utilize it for model training. Typeless also offers improved sentence suggestions for users who may have misphrased a line.

The application permits dictation of up to 4,000 words each week (approximately 16,000 words monthly) under its free tier. A subscription can be purchased for $12 monthly (billed annually) to unlock unlimited words and access new features. Typeless is currently available exclusively for Windows and macOS.

For the past few years, enterprises have been experimenting with various AI tools to determine their strategies for adoption. Investors believe that this phase of exploration is drawing to a close.

TechCrunch recently conducted a survey of 24 venture capitalists focused on enterprises, revealing that a significant majority anticipates an increase in enterprise AI budgets for 2026 — although not universally. Most investors indicated that this budget augmentation will be selective, with many enterprises likely allocating more resources to fewer contracts.

Andrew Ferguson, a vice president at Databricks Ventures, foresees that 2026 will mark a pivotal moment when enterprises begin to consolidate their investments and select frontrunners.

“Currently, enterprises are experimenting with numerous tools for singular use cases, and there’s a surge of startups targeting specific purchasing areas such as [go-to-market], making it challenging to identify distinctions even during [proof of concepts],” Ferguson noted. “As enterprises witness tangible results from AI, they will trim some of the experimentation budget, streamline redundant tools, and reallocate that savings toward the AI technologies that have proven effective.”

Rob Biederman, a managing partner at Asymmetric Capital Partners, concurred. He anticipates that enterprise firms will not only narrow their individual spending but that the overall enterprise landscape will also restrict its AI expenditures to a small number of vendors throughout the industry.

“Budgets will rise for a limited selection of AI products that evidently produce results and will sharply decrease for all others,” Biederman said. “We foresee a bifurcation where a select few vendors capture a disproportionate portion of enterprise AI budgets while many others experience stagnant or declining revenues.”

Targeted investments

Scott Beechuk, a partner at Norwest Venture Partners, believes that enterprises will amplify their spending on tools that enhance the safety of AI for enterprise use.

“Enterprises are now acknowledging that the true investment lies within the safety and oversight measures that ensure AI reliability,” Beechuk stated. “As these features advance and lower risks, organizations will gain the confidence to transition from pilots to broader deployments, leading to an increase in budgets.”

Harsha Kapre, a director at Snowflake Ventures, forecasts that enterprises will allocate funds for AI in three specific areas in 2026: enhancing data foundations, optimizing models post-training, and consolidating tools.

“[Chief investment officers] are actively working to minimize [software-as-a-service] sprawl and are leaning towards unified, intelligent systems that reduce integration costs and provide quantifiable [return on investment],” Kapre remarked. “AI-enabled solutions are likely to benefit most from this transition.”

A transition away from experimentation and towards consolidation will impact startups. However, the precise nature of this effect remains unclear.

AI startups might find themselves facing a reckoning similar to that which SaaS startups experienced a few years back.

Companies with unique, hard-to-duplicate products such as niche solutions or those based on proprietary data will probably continue to thrive. In contrast, startups offering products comparable to those of major enterprise suppliers, like AWS or Salesforce, might begin to see a reduction in pilot projects and funding availability.

Investors are also aware of this potential outcome. When asked how they can identify an AI startup with a competitive edge, several VCs noted that firms boasting proprietary data and products not easily replicable by a tech giant or large language model company are the most secure.

If the investor forecasts prove accurate and enterprises do start to optimize their AI spending next year, 2026 could be a year of increased enterprise budgets, yet many AI startups may not receive a larger share of the market.

Annually, TechCrunch’s Startup Battlefield pitch contest attracts thousands of applicants. We narrow down those submissions to the top 200 candidates, and from those, the top 20 vie on the grand stage to claim victory, claiming the Startup Battlefield Cup and a monetary award of $100,000. However, the other 180 startups also impressed us greatly in their respective sectors and compete in their own pitch contests.

Below is the complete list of the consumer/edtech Startup Battlefield 200 nominees, including a note on their inclusion in the competition. 

Ahoi 

What it does: Assists individuals in locating venues that accommodate those with limited mobility.  

Why it’s noteworthy: Their inclusive technology enhances accessibility for individuals who may find it challenging to locate suitable locations for their needs.  

AllFocal Optics Limited 

What it does: Implements nanophotonic technology to develop lenses that improve visual sharpness.  

Why it’s noteworthy: The firm claims it has developed groundbreaking technology to assist people, particularly those suffering from headaches and dizziness, in enduring extended reality experiences.  

Billight 

What it does: Billight is a light-emitting pool table.  

Why it’s noteworthy: It claims to be the first illuminated pool table and gaming system.  

Cerca Dating  

What it does: This Gen Z dating application is reviving the traditional meeting method — through mutual acquaintances.  

Why it’s noteworthy: It represents the latest take on a dating app amid a period of burnout from romance apps, aiming to showcase that digital love still exists.  

FounderWay.ai 

What it does: A platform that enables startups to grow by offering business guidance on subjects such as crafting a pitch deck or identifying target markets.  

Why it’s noteworthy: The platform utilizes AI to provide answers to some of the most critical questions founders often face — how to effectively operate and expand a business. It offers a more straightforward approach than searching for information across various sources individually.  

Hotel Treats 

What it does: A platform that enables luxury hotels to supply customers with vouchers for amenities like spas and dining.  

Why it’s noteworthy: The platform permits hotels to generate revenue from day passes while simultaneously offering consumers an opportunity to indulge in a unique luxury experience, all without needing to pay substantial amounts and stay overnight at the hotel.  

Jotto 

What it does: A business that develops QR codes for events and venues, allowing attendees to provide feedback and reviews.  

Why it’s noteworthy: It’s a fascinating product that also provides a method for individuals to give feedback via video or voice recording.  

Nim 

What it does: A platform that permits users to generate AI-driven videos.  

Why it’s noteworthy: It’s part of the wave of AI video startups emerging, but it stands out by providing a comprehensive service, including prompt support and a plethora of reusable clips.  

Perfingo 

What it does: Perfingo serves as a financial planning instrument.&nbsp.

Why it’s noteworthy: It proclaims to be the first of its kind in its base location of Singapore.  

Pintours 

What it does: Pintours functions as a tour-booking platform.  

Why it’s noteworthy: It acts as an AI tour guide, enabling users to navigate a tour independently and customize the experience as they choose.  

Prickly Pear Health 

What it does: Prickly Pear offers a voice AI companion for women that oversees brain health.&nbsp.

Why it’s noteworthy: This isn’t a mere chatbot, but an AI that is trained to interpret shifts in language and context which may signal cognitive challenges, particularly those induced by hormonal alterations experienced by women in their 30s to 50s. 

rax 

What it does: Rax serves as a peer-to-peer clothing rental service.  

Why it’s noteworthy: The victor of the premier consumer pitch, Rax claims to be among the first to launch in Canada and has recently announced a venture into the U.S. market.  

Rent a Cyber Friend

What it does: Facilitates connections between individuals seeking friends in their professional fields online.&nbsp.

Why it’s noteworthy: Differentiating from a typical social network, this application aids users in discovering potential friendships and includes functionalities such as video calls and chats to foster new connections. 

Renude 

What it does: Renude provides an AI-driven skin care recommendation platform for beauty brands.&nbsp.

Why it’s noteworthy: Utilizing computer vision AI and LLMs, this e-commerce solution enables skin care companies to deliver personalized product suggestions to each client.&nbsp.

Snap Discovery AG 

What it does: Presents a brain-computer interface designed for hands-free interactions in daily life.&nbsp.

Why it’s noteworthy: Snap integrates with the game development platform Unity and is suited for various applications, ranging from gaming to stress management.&nbsp.

Tasteit 

What it does: Tasteit is an application that assists individuals in finding dining companions.&nbsp.

Why it’s noteworthy: Tasteit promotes itself as the anti-dating app, emphasizing its mission to utilize food and restaurant outings as a means for people to connect and meet.&nbsp.

Tattd 

What it does: Tattd is an AI-centric app facilitating the discovery and booking of tattoo artists.&nbsp.

Why it’s noteworthy: The startup employs generative AI to produce a design mock-up that is subsequently matched with a tattoo artist whose style corresponds with that mock-up.&nbsp.

Vista InnoTech Limited 

What it does: Vista InnoTech has developed technology that enhances photos by mitigating the impacts of accidental movement or unstable environments.&nbsp.

Why it’s noteworthy: They’ve introduced a device known as the Micro Gimbal Stabilizer, compact enough for incorporation into most mobile gadgets, that functions effectively even in dim lighting conditions.&nbsp.

Young Minds App 

What it does: An app for parental controls that monitors and prevents children from partaking in unsafe online activities.&nbsp.

Why it’s noteworthy: The application encourages children’s wise online decisions and features a distraction-blocking option during study periods.&nbsp.

ZoraSafe 

What it does: ZoraSafe recognizes and shields users from scams.&nbsp.

Why it’s noteworthy: Targeting families and elderly individuals, ZoraSafe analyzes links and messages to guard against scams, including those involving deepfakes and social manipulation. It further provides features like AI coaching.&nbsp.

Edtech

Calificadas 

What it does: AI-enabled training to enhance employee communication skills in the workplace.&nbsp.

Why it’s noteworthy: This app for professional growth was crafted with diversity, equity, and inclusion at its core and leverages AI to assist users in refining their word choices, message formats, and even their nonverbal communication styles.&nbsp.

CampusAI 

What it does: CampusAI provides a versatile platform focused on training users about AI.

Why it’s noteworthy: The platform is designed to benefit everyday users looking to implement AI to enhance their work across various fields, including sales, HR, legal, and more.  

General Neuro 

What it does: The NeuroLingo headset aids individuals in acquiring a foreign language.&nbsp.

Why it’s noteworthy: This headset creates an environment that facilitates language acquisition in conjunction with a matching app.&nbsp.

Readmio 

What it does: A storytime application designed for parents and children.&nbsp.

Why it’s noteworthy: The app follows the text as it is read aloud, automatically incorporating sounds and music at specific parts, making the stories more engaging.&nbsp.

Super Teacher 

What it does: Super Teacher provides an AI-based tutor for elementary education.&nbsp.

Why it’s noteworthy: This AI tutor delivers customized instruction and assessments for classroom usage, with round-the-clock access for students at home.&nbsp.

ZEZEDU Corp. 

What it does: Zezedu is an AI-enabled solution, crafted in South Korea, that provides tailored math education.  

Why it’s noteworthy: A mathematics instructional tool for educational institutions and institutes that monitors assignments, grading, and feedback while offering individualized curriculum.&nbsp.