Techin' Geek

A legislative proposal that would permit autonomous vehicles to function in Washington, D.C. has turned into a pivotal point for Uber’s wider robotaxi agenda. Rather than just collaborating with and investing in robotaxi creators, Uber is also seeking to influence the regulations governing them, a strategy that places it at odds with its business associate, Waymo.

Uber, which stands against the bill, contends that the suggested regulation would replace for-hire human drivers and grant Waymo a de facto monopoly. Instead, it has advocated for a framework that would mandate robotaxis to operate within a ride-hailing network that also includes human drivers, based on public records accessed by TechCrunch and discussions with industry and company insiders. 

“We have already observed in various regions how a misguided, exclusive first-party regulatory framework can disrupt an urban area,” Javi Correoso, head of U.S. policy and federal affairs at Uber, remarked in May during a D.C. Council roundtable concerning a different, existing law regulating for-hire drivers. Correoso asserted that robotaxis generate congestion through idling or traveling empty, are unable to offer the physical support that older or disabled individuals need, and referenced statistics indicating that one AV replaces approximately four human drivers.

When inquired about the hybrid model, Correoso articulated Uber’s regulatory perspective. 

“A hybrid model signifies that consumers should have the option to access both. If a consumer uses the app, they should be able to select,” he stated, according to a publicly accessible recording and transcript. “I would suggest going further: this should be embedded in the regulatory framework for the sector. There ought to be a stipulation for users to have access to a human-driven Uber.”

Waymo, owned by Alphabet, argues that the bill it supports will facilitate the secure deployment of autonomous vehicles while enhancing public transportation, equitable access, and support for workers, without curbing businesses like Uber. 

On Monday, the two companies will express their viewpoints during a lengthy hearing. The bill’s approval is not imminent—various stakeholders informed TechCrunch that they hope legislation is enacted before the year concludes and prior to Washington D.C. Mayor Muriel Bowser departing office in January. Nonetheless, the discussions and lobbying efforts surrounding the bill illustrate a larger debate that extends beyond Washington D.C.

The proposed AV bill

Introduced by Councilmember Charles Allen in May, the bill aims to update the Autonomous Vehicle Act of 2012 to allow for driverless testing and commercial operations within the district. At present, companies like Waymo and Zoox conduct autonomous vehicle testing, but only with a human safety operator in the driver’s seat.

The proposed legislation would empower the District Department of Transportation (DDOT) to grant permits for driverless testing and deployment to AV developers that fulfill specific criteria. Such criteria include maintaining a minimum of $5 million in liability coverage and committing to report crash data within either 8 hours or 72 hours, based on whether the vehicle is part of a commercial fleet or a privately owned AV (which is not currently available in the market).

Additionally, the bill would impose a $0.15 per mile tax on robotaxi operators, a suggestion that robotaxi proponents have labeled as overly burdensome. Revenue from this “vehicles miles traveled” (VMT) tax would be divided, with 50% allocated to public transit and the remainder utilized to support education and workforce initiatives for ride-share and taxi drivers at risk of losing their livelihoods to robot vehicles.

The bill has attracted a variety of stakeholders beyond Uber and Waymo. Numerous organizations and companies—including representatives from Tesla, Lyft, the Teamsters and Service Employees International Union labor groups, disability rights and accessibility organizations, local businesses and industry representatives, highway safety advocates, government officials, and think tanks—are all set to contribute to Monday’s hearing. 

The proposal has spurred an anti-robotaxi campaign initiated by a New York-based entity named Coalition for Accountability and Road Safety, which is engaging voters and utilizing social media outreach. The funding sources for the organization remain ambiguous, as it is registered to an employee of Pitta Bishop & Del Giorno LLC, a New York lobbying and government affairs firm associated with labor and employment law firm Pitta LLP.

Publicly available lobbying records show Pitta has been engaged over the past year by numerous labor unions and the New York Black Car Operators’ Injury Compensation Fund. 

The stakes are significant for all robotaxi developers, human drivers, and the ride-hailing and taxi firms that employ them in D.C. The stakes are arguably even higher for Uber and Waymo, given their substantial market influence. Uber is the leading ride-hailing and delivery network in the United States, while Waymo is the top robotaxi provider, offering over 500,000 rides weekly across 11 cities. 

Should Uber succeed and its hybrid network concept be embraced in D.C. — or elsewhere — it would leave AV developers such as Waymo with two options: place their robotaxis on ride-hailing apps like Uber’s or utilize human drivers who offer ride-hailing services alongside robot cars that have taken extensive time and investment to develop.

If Waymo and other advocates of the D.C. bill achieve success, Uber contends it will be relegated entirely.

Protect and expand

Although the bill concerns local policy, it underscores one aspect of Uber’s strategy to safeguard its dominant position in the ride-hailing and delivery sectors. 

Uber is presently investing in and collaborating with over 30 autonomous vehicle technology companies across the globe while also establishing AV Labs, a newly created division aimed at gathering and sharing real-world driving data with AV creators. The firm is recruiting numerous engineers for this division, according to job postings and discussions with knowledgeable sources. 

As Uber asserts its presence in the AV landscape, it is simultaneously promoting protective regulations that mandate autonomous vehicles to function alongside human drivers within a unified platform — much like the Uber app. 

Uber’s investment and partnership initiatives have been ongoing for several years. The company’s advocacy for a hybrid network is a more recent development, first appearing in a white paper released in May. Since then, Uber has increased its advocacy with policymakers, including the D.C. Council roundtable meeting in May aimed at revising the district’s Vehicle-for-Hire Innovation Amendment Act of 2014. (This legislation, which manages ride-hailing and taxi services through the Department of For-Hire Vehicles, is distinct from the AV bill, but multiple sources have informed TechCrunch that the policies overlap.)

In June, Uber sent a letter to the D.C. Council, which TechCrunch has reviewed, further elaborating on Correoso’s prior statements. The letter noted that the hybrid concept would represent a singular transportation network with traditional drivers that subsequently incorporates autonomous vehicles.

“In practical terms, this means that if you request an Uber in an area with AVs, you may be paired with either an AV or a human driver, depending on the specifics of your journey,” the letter indicated. 

In D.C., Uber is reacting to a proposal that would essentially prohibit hybrid networks, as stated by company spokesperson Noah Edwardsen to TechCrunch. 

Waymo contests this interpretation, with a company representative stating that Waymo does not endorse attempts to restrict AVs to particular types of networks. “We would welcome modifications that make it clear that different network types can function in the District,” wrote Waymo spokesperson Ethan Teicher in an email to TechCrunch.

More broadly, Edwardsen claimed that Uber has not adopted a one-size-fits-all regulatory approach, contrasting it with “advocacy from segments of the AV sector today, where proposals have frequently failed to address significant issues like labor and transportation equity — or that have attempted to opportunistically exclude competitors and create monopolies — making them mostly unviable.”

While a variety of industry experts have critiqued certain elements of the D.C. bill — particularly the VMT tax and proposed robotaxi cap — some are opposed to Uber’s hybrid proposition. 

Greg Rogers, founder and executive director of the nonprofit mobility and tech think tank The Innovation Majority, is expected to speak at Monday’s hearing and characterized Uber’s maneuver as an effort at “regulatory capture.”

“Mobility is already an existing marketplace — individuals can currently choose whether to use a bus, ride a bike, walk, or take a rideshare each day,” Rogers told TechCrunch in an interview. “And any argument suggesting that consumer welfare can be enhanced by mandating specific business models while eliminating others does not enhance people’s mobility options. It does not improve road safety, and instead risks further entrenching interests and monetizing anyone seeking to operate AVs in the district.”

Uber’s pro-driver, “let’s find common ground” stance may astonish keen observers of the ride-hailing entity. The company’s formative years were characterized by an anti-regulation mindset that sought loopholes within current laws or altogether disregarded them.

Uber has frequently opposed regulations supported by labor unions, such as AB 5 in California, which would have disrupted its asset-light business model by classifying gig workers as employees. Proposition 22, a 2020 ballot measure passed by voters and validated by the California Supreme Court, was supported by Uber, Lyft, and others as a compromise that afforded workers access to health insurance and other benefits while preserving their contractor status.

These confrontations, along with others of a similar nature, have taught Uber the necessity of considering human employees and the influence of labor unions that advocate for them, should it aspire to occupy a central role in the robotaxi market, according to sources. Uber’s own chief operating officer Andrew MacDonald adopted a similar tone of having “learned from past experiences” in a LinkedIn post in May that promoted its white paper.

MacDonald remarked that the repercussions of the company’s aggressive growth tactics led to “regulatory conflicts and a corporate crisis that undermined trust for years.”

“That experience transformed us,” he asserted. “Nowadays, we collaborate with municipalities instead of confronting them.”

Uber contends that its hybrid network proposal represents that compromise — allowing for the coexistence of robotaxis and human drivers on the same platform while mitigating labor concerns.

The company is dedicated to advocating for this idea in other cities and states as legislators formulate new AV regulations or amend existing ones.

Wired released its own report elaborating on lobbying endeavors in New Jersey and D.C.

Uber’s position, coupled with its active lobbying, places it on a collision path with Waymo. 

Frenemies

Waymo and Uber have previously clashed over autonomous vehicle innovation.

In 2017, Waymo initiated a lawsuit against Uber over accusations of trade secret misappropriation. The highly publicized court case, where Waymo alleged that Uber used trade secrets illicitly obtained by former Google engineer Anthony Levandowski, featured memorable testimony and revealed phrases like “laser is the sauce.” The trial concluded in just five days when Uber opted for a settlement, leading the two companies to cease their public disputes.

Fast forward six years, with Uber’s internal AV development program divested to Aurora, the former adversaries joined forces. Waymo agreed to integrate its self-driving cars into Uber’s app in Phoenix in 2023. However, this partnership quietly concluded in May, characterized as limited and a “pilot.” Waymo also maintains its standalone app in Phoenix, its inaugural robotaxi market.

The relationship appeared to strengthen by March 2025, when company leaders, enjoying prickly pear margaritas and plates of Terry Black’s barbecue at a private gathering, celebrated the introduction of Waymo robotaxis on the Uber app in Austin during the annual SXSW festival. This partnership soon expanded to Atlanta. In both cities, eager customers cannot directly summon a robotaxi through Waymo’s app; they must rely on the Uber app and hope for a successful match.

Recently, however, the relationship has taken a sour turn—publicly.

Earlier this year, Uber chief technology officer Praveen Neppalli openly criticized Waymo on X, sharing a video and commentary highlighting the unsafe and “scary” actions of a Waymo robotaxi. During an earnings call in May, Uber CEO Dara Khosrowshahi made remarks directed at Waymo without naming the company, expressing support for regulatory oversight.

“They’re posing the right inquiries, which concern how AVs will manage interactions in situations with power outages or in school zones or collaborating with firefighters, etc., in urban environments,” Khosrowshahi stated, referencing recent incidents involving Waymo robotaxis. 

The tension between Waymo and Uber has even reached an international level, as both companies prepare for an anticipated showdown in London.

Amid rising speculation about the timeline for potential collapses of Uber and Waymo’s existing collaborations in Austin or Atlanta, both entities are preparing for a regulatory battle that seems likely to extend into other cities and states.

Uber is betting, and lobbying for, a distinct future than the one Waymo envisions.

“We believe that the future of our transportation system will be hybrid,” asserted Uber’s head of AV policy Harry Hartfield in testimony given prior to Monday’s meeting. “Public policy should be structured around that reality, rather than an AV-exclusive scenario that does not exist.”

Glad to see you back on TechCrunch Mobility, your go-to for innovations in transportation and, increasingly, the role of AI. To receive this directly in your inbox, sign up for free — simply click TechCrunch Mobility!

I’ve returned from my break. What news did I miss? Quite a lot, actually — notably, the conclusion of the Uber-Waymo alliance in Phoenix. However, Uber and Waymo continue to offer robotaxi services in Atlanta and Austin. The real question isn’t if, but when will these associations come to an end? Yet, I find the more fascinating inquiry to be how the two firms will respond once the remaining partnerships dissolve. 

Tensions are already apparent, with Uber leaders making subtle jabs at Waymo. I anticipate that when the partnerships cease, these thinly disguised insults will evolve into more overt actions. One area of contention will be policy, focusing on the markets where robotaxi companies are seeking entry. 

This week brought another noteworthy twist in the self-driving vehicle landscape at the federal level. National Highway Traffic Safety Administration administrator Jonathan Morrison issued a message to autonomous vehicle manufacturers, declaring it intolerable for their cars to obstruct first responders or police forces.

The key statement: “To be clear: the failure to recognize and respond suitably to these situations represents a significant shortcoming. Emergency scenarios are not unusual or extreme ‘edge cases.’ Consequently, NHTSA is today calling on AV developers and operators to urgently direct their efforts towards addressing this issue.”

Morrison’s letter refrains from naming any specific robotaxi company, as it was dispatched to all AV developers specified in the Department of Transportation’s Standing General Order. However, it certainly appears that Morrison is targeting Waymo’s operations.

An earlier TechCrunch investigation revealed that Waymo — which manages the largest robotaxi fleet in the U.S., with vehicles operating in cities like Los Angeles, Phoenix, and San Francisco — has had numerous encounters with first responders. This week, San Francisco supervisor Bilal Mahmood announced intentions to submit an inquiry letter to investigate how autonomous vehicles impacted public transport and emergency services after a July 4th fireworks event that led to severe traffic congestion. Local media noted that several Waymo robotaxis had to be towed after exhausting their batteries during the long traffic pile-up.

Morrison’s letter carries weight. But will it yield real repercussions for AV developers? It’s difficult to say at this juncture. For now, the NHTSA has mandated that companies present the agency with “solutions” by the month’s end.

Another piece of news from federal authorities. Check out the new 2026 Regulatory Plan and Unified Agenda, which was refreshed last week. It includes a lengthy list of proposed modifications to Federal Motor Vehicle Safety Standards (FMVSS), covering vehicle design and equipment stipulations. These proposed alterations could benefit autonomous vehicle firms like Tesla and Zoox, which are working on vehicles lacking steering wheels, pedals, or other features expected on traditionally-manipulated cars.

A little bird

Have a tip for us? Reach out to Kirsten Korosec at [email protected] or via Signal at kkorosec.07, or contact Sean O’Kane at [email protected].

Deals!

While we typically focus on venture transactions, this week I wanted to shine a light on Rivian and its sale of 86.25 million Class A common shares priced at $15.50 each (this includes an extra 11.25 million in additional shares that underwriters opted to purchase).

In total, Rivian anticipates raising $1.32 billion in fresh capital. This fundraising comes at a significant moment for the electric vehicle manufacturer. The company recently started delivering its new R2 SUV and has increased its sales forecast for 2026. Rivian now expects to deliver between 65,000 and 70,000 vehicles after exceeding its own projections in the second quarter due to strong quarter-over-quarter growth in EDV and R1, along with the arrival of R2 deliveries. 

The company did not clarify the reason for this capital raise. However, it’s worth noting that Rivian is not yet profitable, and ramping up production for the R2 — or any vehicle, for that matter — is costly!

Other deals that caught my eye …

Bidbus, a startup from Los Angeles that has developed a digital marketplace for multiple dealers to submit bids on a car, raised $15 million in a Series A funding round led by Ibex Investors. Mucker Capital, FJ Labs, Motley Fool Ventures, Data Point Capital, Walter Ventures, and the Car Dealership Guy’s Yossi Levi also contributed.

Lyft announced plans to acquire Serveo’s bike-share division in Spain. The terms were not disclosed, but the ride-hailing firm stated it expects to finalize the deal this year.

TaiSan, a battery startup based in the U.K., secured £4.65 million in a seed funding round co-led by Eos Advisory and the Midlands Engine Investment Fund II. Participants also included InnoEnergy, AFI Ventures, EverQuest Capital Partners, Exergon, Heartfelt Ventures, Adeline Arts & Science, Techmind, angel investor François Badelon, and matched funding from Innovate UK.

Notable reads and other tidbits

AssuranceAmerica, a U.S.-based insurance firm, confirmed a data breach affecting the personal information and driver’s license numbers of 6.9 million individuals, marking it as the largest known leak of Americans’ driver’s license data this year.

Beta Technologies, a developer of electric vehicles capable of takeoff and landing, completed operational flights conducted under the new eVTOL Integration Pilot Program from the U.S. Department of Transportation and Federal Aviation Administration. The flights covered approximately 275 nautical miles across Virginia and Maryland. 

Longtime observers of Tesla will recall the exciting times when Elon Musk faced off against various short sellers of the company’s stock. Musk remains a divisive figure, and one creator of an exchange-traded fund has found a means to capitalize on that negative sentiment with two innovative anti-Elon exchange-traded funds. 

GM brand Chevrolet developed a fully American EV truck. Senior reporter Tim De Chant questions, Why is there such low demand for it? 

Manna Aero, the Dublin-based autonomous drone delivery startup, is expanding its presence in the U.S. with a manufacturing plant and operations center in Tulsa, Oklahoma, which it states will create 1,000 jobs in the coming years. 

Slate Auto partnered with Crayola to provide vehicle wraps in five crayon colors for customers of its EV truck and SUV. (Note: The standard Slate EV vehicle isn’t painted. Instead, it features a gray composite material that can be customized with a vehicle wrap. The company offers numerous design choices.)

One more thing …

TechCrunch podcast Build Mode has officially kicked off its third season, and it’s a hit. Build Mode is hosted by Isabelle Johannessen, who leads TechCrunch’s Startup Battlefield program. Unlike Equity — the TC podcast that I co-host with Anthony Ha and Sean O’Kane — Build Mode aims to assist early-stage founders. 

The new season begins with Precursor Ventures founder and managing partner Charles Hudson, who discusses what early-stage entrepreneurs should know before seeking their initial institutional funding.

Be sure to check it out: The new strategies for early-stage fundraising with Charles Hudson.

Reed Jobs is quite personable. He talks a lot, is self-effacing, often makes video-game references, and has a genuine passion for his work. While he isn’t eager to dwell on being the son of Steve Jobs, he also doesn’t shy away from it. When our producer, Maggie, inquired if he was using a MacBook during our video call on Thursday morning, he responded instantly: “Are you serious?”

What he prefers to discuss is Yosemite, the oncology-centric venture firm he founded in 2023 to, among other things, develop biotech firms from the ground up based on initial academic research, combining philanthropy with external investment capital. Three years later, Jobs is determined to elevate Yosemite to a major player, not merely for the sake of competition but because he believes the prospects ahead of him are growing more rapidly than he anticipated, thanks to AI influencing both drug discovery and clinical trial methodologies.

Companies in his portfolio that he takes the most pride in include Azalea, which originated from a grant to Jennifer Doudna’s lab and is now in clinical trials, and Quarry, a venture co-founded with serial entrepreneur Craig Crews that revolves around a novel therapeutic method known as induced proximity, where a drug functions by physically adjacent a disease-inducing protein to the cell’s own degradation system (instead of attempting to obstruct it directly).

The last time we connected with Jobs at TechCrunch Disrupt nearly three years ago, Yosemite was freshly launched, and the biotech sector was still reeling from its post-pandemic downturn. Now, the firm employs 17 people; a wave of significant drugs are nearing the end of their patent protections almost simultaneously, creating numerous new opportunities; and AI has transitioned from a mere curiosity to, in Jobs’s own words, a central component of Yosemite’s operations. We discussed all of this.

This Q&A has been shortened for brevity.

TC: Earlier this year, you announced the initial close of your second fund aiming for $350 million. What’s the current status at Yosemite?

RJ: It’s a period of intense activity for us. We’ve experienced remarkable traction and onboarded many significant new partners. Yosemite stands out as a venture entity for two main reasons: we focus exclusively on oncology — that represents 40% of the biotech landscape — and we aim to create our own companies. We’re convinced that cancer cures aren’t just waiting to be discovered in pharma; we need to innovatively create them using new insights. To mitigate risks on those concepts early on, while they’re still nascent in university labs, we apply some philanthropy without any strings attached. Two of the 20 companies in our first fund arose directly from a grant.

What portion of that $350 million is allocated to companies you’re establishing versus those you’re participating in?

Approximately a third is dedicated to companies we’re creating ourselves — either based on our own ideas or in collaboration with academics at institutions like Yale, Berkeley, and Stanford. This process requires considerable time and effort, which is why it only comprises a third. The remaining amount is directed towards existing companies which we aim to join. Furthermore, 2.5% of the fund’s [assets under management] will be allocated to a donor-advised fund — that’s completely unrestricted grant money, along with $1 million annually from our management fees.

Though it’s early, what arguments do you present to potential LPs about performance compared to other life science VC firms?

We are still in the early stages, but Yosemite is positioned to establish new medical domains ahead of other firms. My team has been at the forefront of a few of these: epigenetic gene editing [technology that modifies how strongly a gene is expressed, without changing the actual DNA sequence], and safe delivery for gene editing to targeted cells — a long-standing challenge in the entire field. If your goal is to be a pioneer and assist in uncovering new zones, that’s where our strengths lie.

Previously, you expressed concerns regarding the cautious nature of biotech investors. Has that situation changed?

It has indeed. When I established Yosemite in 2023, the XBI [ETF/index] was still significantly down from its peaks in 2021, and pharmaceutical companies hadn’t begun making acquisitions yet. The last three years have seen: improving interest rates and pharma facing its most significant patent cliff in history while holding substantial cash reserves from the pandemic. This has culminated in an acquisition surge over the past eight months. We’ve seen remarkable exits, such as Eli Lilly purchasing Kelonia for $7 billion, and significant victories in antibody-drug conjugates. A notable example: Revolution Medicines, addressing KRAS [one of the most frequently mutated cancer-driving genes, which was long deemed nearly impossible to target with drugs] in pancreatic cancer, has doubled the survival rate for [the most prevalent form of pancreatic cancer] — from 12 to 24 months. This progress has all been achieved within the last year.

Last year, you publicly voiced concerns over proposed NIH budget cuts.

Regrettably, federal government pressure remains, though it’s less of a long-term threat than it once was. Last year, historically, an administration sought cuts of up to 40% to the NIH budget. For reference, the largest previous cut was 1% in 2009, due to the global financial crisis, which resulted in 7,000 NIH scientists losing their jobs. Luckily, both the Senate and House — this is highly bipartisan — completely rejected the 40% cut. This year, they requested 12%, still the largest cut ever proposed by a significant margin, and I anticipate a similar dismissal. NIH funding enjoys over 90% approval. Personally, I advocate for an offensive approach — I’d boost it to around $100 billion. In dollar terms, it hasn’t grown in about ten years, so adjusted for inflation, it has actually diminished.

In which areas is AI currently influencing healthcare delivery?

American hospitals are among the most technologically backward sectors in the economy — a lot still relies on fax machines and floppy disks. An example is the call centers, akin to 911 triage, which are costly to maintain round-the-clock and are ripe for AI utilization. There’s also the realm of electronic health records, radiology, and pathology. However, my keen interest is in clinical trials — these represent the largest financial burden and time consumption in drug development. A Phase 3 cancer trial incurs about $260 million, with only one-third achieving success. The largest expense is patient recruitment and retention. AI could aid in constructing a synthetic control arm [a computer-generated substitute for the untreated comparison group, based on existing patient data], eliminating the need for a complete control group and potentially halving the required patient count while greatly enhancing speed. The FDA is currently encouraging this direction.

How about AI in drug discovery — is it overstated?

I believe it’s a terrific breakthrough, encouraging democratization of science and speeding processes up. Currently, AI accelerates much of the tedious work — not necessarily improving quality, but executing it at remarkable speeds with consistent results.

AI has [also] been effective in identifying areas we’ve previously struggled to access. Traditionally, we could only target about 15% of the genome because we faced challenges with proteins that interact with one another — the chemistry was too complex. This landscape has transformed over the past few years, closely following AI advancements. Consider Revolution Medicines: they pioneered the ability to target KRAS, which for decades lacked any [natural indentation for a drug molecule to attach and inhibit] — it’s essentially a smooth oval, akin to a death star. Roughly a decade ago, Amgen scientists discovered a peculiar hidden pocket within it, leading to the initial drug against it, Lumakras. This was effective only for one specific mutation; however, AI has since identified all the other variants we can now target and proposed innovative methods to inhibit it.

Which undruggable targets are your companies pursuing?

The largest target is p53. We’re tackling it with three companies employing different approaches. It’s a tumor suppressor gene — notably, elephants are largely cancer-free, and one hypothesis suggests they have many copies of p53, whereas humans possess only a single copy, making it easier to eliminate. p53 is the most commonly suppressed gene across human cancers; nearly all cancers must disable it to become viable initially. If we could reactivate it, or focus on its mutated forms, this might present a critical weakness in cancer, a feat that has yet to be achieved. We believe we’ve identified a way to target that exposed [marker] across the various mutations of p53.

Can you elaborate on Tune Therapeutics?

Tune has excelled as the leading epigenetic editing firm in clinical development for the past few years, focusing on hepatitis B, which impacts over 250 million individuals and is a primary contributor to liver cancer. This technology enables us to add or eliminate methyl groups [tiny chemical tags that attach to DNA and operate like dimmer switches, modulating a gene’s activity without altering the gene itself] at designated locations in the liver. Every cell within your body shares the same DNA but expresses it differently — consider gray hair: when melanin becomes methylated and deactivated, your body still produces hair, albeit less vibrantly. This mechanism is also responsible for aging immune systems and decreased metabolism. Hepatitis B appears foreign to your immune system, so we aim to methylate and deactivate the virus, mirroring the natural response of about 1% of individuals who spontaneously clear the virus.

Additionally, Histosonics is a device company, which seems atypical for Yosemite.

Correct, we don’t typically focus on devices. It’s the first firm employing histotripsy on a large scale for the destruction of liver tumors, utilizing noninvasive methods — it creates small air pockets and then collapses them to damage tissue in a very targeted manner, likened to an ultrasound rather than a CT scan. Their primary initiatives involve pancreatic and liver tumors — since most pancreatic cancer spreads to the liver, it’s a logical match. We anticipate this will play a significant role in treatment for both conditions.

What’s the current number of companies in your portfolio, and have there been any failures thus far?

Nearly 25 across both funds. Two have not succeeded due to scientific reasons — we evaluate these investments against scientific benchmarks, and given our early stage, sometimes failures occur on the scientific front. That’s to be expected.

What guidance do you offer founders considering a substantial investment from large pharmaceutical companies? You secure the funding, but it limits other alternatives.

Pharmaceutical companies are essential partners, but founders should view them as a variable — priorities can shift significantly with leadership changes. Following COVID, many pharma corporations, like Pfizer, incurred losses in infectious disease and exited the field entirely. Staying updated on who is actively involved in your domain is likely the most crucial aspect.

How can aspiring founders engage with you?

We maintain an open-door policy. When reviewing grants and companies, we remove any identifiers from people’s CVs — I prefer not to know whose idea it is or what position someone holds. We’ve supported labs led by Nobel laureates and those with no prior grant experience equally. We consider all modalities — small molecules, radiopharmaceuticals, gene therapy, immunotherapy, AI, digital health. Please reach out via email. Any concept capable of impacting cancer patients interests us.

Is storytelling as vital for biotech founders as it is in other sectors?

Regrettably, it is — I’ve observed companies with excellent science fail due to poor storytelling from their CEOs. However, typically, the founder and CEO are not the same individual. Founders often come from academia, serving as the chief scientist or chief medical officer, while the CEO is a professional operator responsible for fundraising and narrative construction. This division of roles tends to be effective.

Three years into leading Yosemite, what’s been your most significant surprise?

We now have our first trillion-dollar pharmaceutical company, Eli Lilly, due to GLP-1s — the top-selling drug category globally. We’re also observing initial indications that GLP-1s may offer protection against neurodegenerative diseases and cancer, beyond merely facilitating weight loss, as obesity is one of the two leading “pan-disease” risk factors — the other being smoking — that elevate the risk across nearly all disease categories. This has sparked renewed interest, ambition, and investment in vast disease territories that had become stagnant. Genes such as KRAS, Myc, beta-catenin, and p53 — the pantheon of oncogenes that have eluded us for years — are now, in our view, attainable. I didn’t foresee Yosemite progressing at this pace. This moment is more crucial than I initially realized, making it both more daunting and more empowering.

Before we conclude, what are your thoughts on the longevity sector?

I’m not ready to die anytime soon, and longevity is a personal priority. However, I believe we — like everyone else — lack a comprehensive understanding in this area. Ask a geneticist, and they’ll mention telomeres; consult an immunologist, and they’ll discuss T cells losing efficacy; a metabolomicist will have a completely different perspective. There isn’t a unified theory of aging analogous to what we have in physics. I don’t think there’s a singular “longevity problem”; rather, human bodies age differently across various cell types, and the interactions among these parts constitute what we term aging. Personalizing that optimization for each individual is precisely what healthcare ought to pursue, but I’m uncertain how to translate longevity into a one-size-fits-all commercial venture.

The harsh heat wave in NYC last weekend left me desiring a frozen beverage almost every day. Usually, that would mean enduring a trek to 7-Eleven for a slurpee. However, this time, I opted to stay in and put the new Ninja Slushi Twist through its paces.

Ninja’s latest offering in the slushie category builds on the success of the original Slushi but boasts a significant upgrade. Instead of having a single mixing chamber, the Slushi Twist features two 48-ounce containers that can produce two entirely distinct frozen beverages simultaneously. In addition to traditional slushies, it is capable of making frappés, milkshakes, frozen coffees, and smoothies — and since both sides operate independently, you can keep one side alcohol-free for mocktails while the other creates cocktails.

This machine truly addresses a key issue for entertaining. Rather than preparing several batches or getting everyone to agree on one flavor, both compartments operate at the same time, creating enough frozen beverages for larger groups. Ninja claims it can produce over 10 drinks per batch — more than enough for summer get-togethers or family reunions.

The standout attribute is what Ninja refers to as “Dual SlushAssist” technology. Simply put, the device detects what’s inside each chamber and automatically adjusts the freezing temperature for that side. This ensures that a creamy milkshake maintains its smoothness on one side while a fruity slush reaches the ideal icy texture on the other.

However, my favorite feature is the twist dispenser. You have the option to pour each beverage separately or adjust the dial on the drip tray to blend both flavors into a single glass, which is fantastic for creating layered drinks that appear far more sophisticated than the effort required to make them.

Cleaning is straightforward as well. Simply press the rinse button and continue adding warm water until it flows clear.

There are a couple of considerations to think about before making a purchase. The Slushi Twist is noticeably bulkier than the original version, so it’s wise to ensure you have adequate counter or storage space. At $399.99, it also represents a considerable investment. If you only occasionally whip up a frozen drink, this might be more machine than you require. However, for larger families or anyone who enjoys hosting, the dual-vessel design strongly justifies its presence.

Over the last few years, several technology leaders have suggested that eyewear could become the next significant interface for consumer gadgets. However, the current generation of smart glasses still depends greatly on smartphones, despite possessing decent hardware. Even Realities’ G2 smart glasses share this limitation. They present an upscale appearance with a neon-style heads-up display visible in various lighting conditions — yet their performance is largely contingent on their connection to a mobile device, which can be** inconsistent and exasperating.

Differing from competitors like Meta, Even Realities employs a unique approach to smart glasses. The devices feature a monochrome heads-up display that presents text and data in green, reminiscent of a neon sign.

The design intentionally excludes cameras and speakers, focusing instead on productivity and alleviating concerns that those nearby might be recorded.

The G2 represents Even Realities’ second iteration of smart glasses and offers enhancements over the earlier G1 model introduced a few years back. The G2 features a brighter 1,200-nit display (compared to 1,000 nits on the G1), an increase from two to four microphones, and a display area that is 75% larger than its predecessor. Additionally, the new display boasts a superior 60Hz refresh rate, as opposed to the G1’s 20Hz.

Throughout the few months I utilized the G2, the connection to the smartphone has significantly improved. Initially, the glasses frequently disconnected from the app, making me almost give up on using them. However, after a series of app updates, that problem diminished.

These glasses are aimed at individuals who frequently attend meetings, deliver presentations, and travel to multilingual environments. 

Design

Available in two frame styles, the glasses are remarkably light at 35 grams. Constructed from magnesium alloy with titanium alloy temples (the arms that rest over the ears), they are comfortable to wear regarding weight and fit. 

Given that I typically work from home, I found little reason to wear them throughout the day. Nonetheless, the lenses are equipped with UV protection, making them suitable for outdoor wear to shield against eye damage — irrespective of smart features.

The company asserts that, under typical usage, the G2’s battery can last up to two days on a single charge. It comes with a protective case capable of recharging the glasses up to seven times before requiring external power itself. While I did not personally verify the two-day duration claim, the battery life was sufficient for me to recharge them without depleting their charge completely.

That case is quite large — it won’t fit in a pocket — but it is sturdy, and the glasses rest securely inside.

Features and operation

These glasses serve as aids for scheduling, reminders, and accessing notes. You can activate them through taps on the stem-based controls. A double-tap on the control pad brings up a dashboard displaying information such as upcoming meetings, stock updates, and top news. 

The G2 can also display real-time phone notifications, but the pop-ups were not consistently reliable — and since my phone is usually close by, I found limited use for this feature.

Long-pressing the temple control reveals a menu with various functions: a notifications tray, Translate, Conversate, Teleprompt, a to-do list, and Navigate. The Translate feature allows you to designate a target language and communicate with anyone. At the recent Global Connect Show (GCS) in China, I tested the glasses while interacting with company representatives conducting demos, and the translation was sufficient for me to follow along when a speaker used Chinese. I also engaged with fellow journalists conversing in different languages, such as French and Spanish. (A limitation of this feature is that the other person won’t understand your language unless they are using the app as well.)

The Navigate feature is intriguing, providing turn-by-turn directions on the heads-up display. The drawback is that it does not integrate with Google or Apple Maps. Instead, you must plan your route via the Even Realities app. I tested it a few times while walking to cafés near my home. The directions were clearly displayed, but the app misidentified some addresses, making it less reliable for unfamiliar destinations. However, I can envision it being beneficial for cyclists or motorbike riders once the company addresses the accuracy concerns.

[embedded content]

Conversate initially displayed a live transcript of discussions on the glasses, which seemed unnecessary since one could easily record a meeting with an app or a separate notetaker. Subsequently, the company incorporated a “prep notes” feature that introduces contextual relevance: you can input notes or documents prior to a meeting, enabling the AI to reference them during the conversation or listen in real time and provide brief explanations for concepts as they arise. For example, during a session about energy, it displayed a bubble for “Green Hydrogen,” and tapping on it revealed a definition right in my line of sight. That feature was significantly beneficial — though I wouldn’t want a transcript or explanatory bubbles for every discussion I engage in.

Central to all this is the integrated assistant, Even AI. Like any voice assistant, it requires a wake word for activation and allows you to ask questions or add to your to-do list. My instructions for to-do items were frequently misunderstood, and for general inquiries, the responses were often lengthy paragraphs that scrolled across the screen without an option to skip or interrupt.

Another drawback is that, despite featuring four microphones, Even AI often struggled to respond or misinterpreted my commands when I was outdoors. The background noise in India may have contributed to this issue, but I would still expect a modern device to exhibit superior noise management.

The G2’s display was easily readable in most situations, but I had to manually adjust the brightness through the app in particularly bright environments. Even though the company hasn’t integrated an automatic-brightness sensor, I would appreciate a manual brightness control accessible directly on the glasses rather than relying on the phone app.

Don’t consider the R1 ring

Even introduced a companion ring called the R1 alongside the G2, intended to control the glasses via a touch interface on the ring rather than using the glasses’ built-in touch controls. However, its price and functionality fail to make a strong case for its value.

The ring performed well, and I experienced no issues during use. Nonetheless, I found it challenging to identify scenarios where it was truly necessary, given that the touch-sensitive elements on the glasses already fulfill the same function.

Moreover, Even integrated health monitoring into the ring, tracking heart rate, calories, steps, sleep, and SpO2 (blood oxygen levels). Personally, I would prefer a dedicated ring like Oura or Ultrahuman if I wanted that type of health monitoring in that specific format. Additionally, if I already owned a fitness tracker, I wouldn’t want to invest in a ring that makes health tracking an ancillary function for a device designed to control the glasses. 

These diverse functionalities increase the ring’s price to $249, which isn’t inexpensive. If I used my smart glasses frequently, I might consider purchasing a control ring at a lower price, especially if it included a microphone for issuing commands to the AI assistant. Currently, I would forgo the R1. 

Where does Even G2 fit in?

Smart glasses are emerging rapidly. While models equipped with cameras and devoid of screens, such as the Meta Ray-Bans, are trending, competitors like Meta and Snap are hastening to develop glasses featuring color displays as well. Only a select number of Chinese firms — such as Rokid and Inmo — are offering glasses with a similar neon-display design.

Priced at $599, the Even G2 offers commendable hardware in a lightweight, aesthetically pleasing framework. The company is also striving to enhance the glasses’ customizability by introducing support for third-party applications, although I did not encounter any apps compelling enough to encourage more frequent use of the glasses. They serve as a desirable accessory: exciting for those who enjoy experimenting with new technology and don’t mind trying out diverse third-party applications.

The hardware itself is commendable, but outside of tasks demanding continuous translation or teleprompting, discovering a clear daily application for such smart glasses remains challenging.

Even’s philosophy is that eliminating cameras and speakers positions the product as a productivity-oriented tool — and I concur with this strategy. However, now that the company has achieved unicorn status, it must develop more first-party software to ensure the glasses become a daily go-to item for users.

Over three years after the introduction of ChatGPT popularized generative AI, OpenAI is expanding its target audience from individual users to families.

OpenAI is looking for a specialized product manager in San Francisco tasked with creating experiences for families, caregivers, and older adults across its offerings. The position requires a background in developing products aimed at parents and families, as well as other consumer experiences sensitive to trust, based on the job listing.

This recruitment coincides with ChatGPT’s user base expanding beyond just younger individuals. Estimates from Sensor Tower, shared exclusively with TechCrunch, show that the proportion of global ChatGPT users aged 35 and older increased to 31% in Q2, up from 26% a year prior, while users aged 18 to 24 decreased from 34% to 29%. In the United States, nearly one in four smartphone parents utilized ChatGPT during the quarter, rising from 16% in the previous year, according to the firm’s estimates.

OpenAI did not reply to inquiries regarding the job listing.

The creation of a product role centered on families indicates that OpenAI is starting to view its offerings less as tools for solitary productivity and more as technology tailored for households, remarked Ben Bajarin, CEO of the technology consultancy Creative Strategies.

“This mirrors the trajectory that Google, Apple, and Meta eventually followed as their platforms became woven into daily life, but AI elevates the stakes since the assistant not only mediates content or devices,” he stated to TechCrunch.

This transition also presents new challenges related to trust and safety. Stephen Balkam, the chief executive of the Family Online Safety Institute, noted that the hiring reflects both the evolution of OpenAI and an increasing acknowledgment that AI products utilized by minors require distinct protections compared to those created for adults.

“I view this as safety through redesign,” Balkam shared with TechCrunch. “You take the initial product or service that was released… not really considering kids… so this is a much-needed reaction and response.”

These remarks come as new research released this week by the Family Online Safety Institute discovered that parents are underestimating how frequently their children use generative AI. While 27% of U.S. parents indicated their child had used generative AI in the past week, 38% of children reported using it themselves, according to a survey of over 4,000 families in the U.S. and Australia.

Balkam told TechCrunch that AI firms should develop products with younger users in mind, implementing stronger content controls, age-appropriate experiences, parental oversight, and reminders to make it clear users are interacting with an AI — not a human.

The recruitment also arises amidst increasing scrutiny regarding how AI companies safeguard younger users. OpenAI has faced several lawsuits from parents alleging that ChatGPT has contributed to harm experienced by their children, including incidents linked to suicide.

In addressing some of these worries, OpenAI has implemented a range of safety initiatives over the past year, including parental controls for teenage accounts, directing sensitive dialogues to reasoning models intended to better manage signs of distress, and, more recently, an optional “Trusted Contact” feature that can notify a family member or caregiver in potential self-harm situations.

AI companies, according to Balkam, have a chance to learn from the errors made by social media platforms, which for years treated children similarly to adults before enhancing their safeguards under growing public and regulatory pressure.

The hiring also supports OpenAI’s extensive initiatives surrounding families. During a recent workshop in collaboration with the San Antonio Spurs Community Impact organization and the Positive Coaching Alliance, the company expressed its goal to explore AI’s role in education, coaching, and youth engagement.

Nevertheless, the demographic transition is not exclusive to ChatGPT, even though OpenAI’s audience is evolving in some unique ways.

Sensor Tower estimates that individuals aged 25 to 34 make up 40% of the total app user base for Anthropic’s Claude and Google’s Gemini, similar to ChatGPT, contrasting with 33% for Microsoft’s Copilot. Copilot, however, skews older, with 20% of its users aged 45 and up, compared to 14% for Claude, 12% for Gemini, and 11% for ChatGPT.

While ChatGPT remains relatively less popular among older users, its growth rate is surpassing that of its competitors. The percentage of users aged 45 and over climbed by three percentage points year-over-year in the second quarter, contrasting with a two-point rise for Copilot and declines for Claude and Gemini, based on Sensor Tower’s analysis.

Among U.S. smartphone parents, Gemini demonstrated the broadest reach at 32% in Q2, followed by ChatGPT at 24%, Claude at 4%, and Copilot at 2%.

For Bajarin, the decision by OpenAI to recruit a product manager focused on families indicates the future direction of consumer AI. As AI evolves into technology utilized across different generations, he anticipates companies will introduce family plans, child and teen profiles, caregiver tools, shared household memories, AI tutoring, and enhanced safety measures.

The U.S. federal cybersecurity organization CISA stated that it lacked a pre-established response strategy for addressing a cybersecurity event in May, after an investigative journalist informed them that a contractor had inadvertently revealed sensitive keys and access credentials for U.S. government systems.

CISA, the Homeland Security division responsible for protecting federal networks and assisting in the defense of critical infrastructure, disclosed on Friday in a post-incident report that its personnel “had to allocate time to develop [a playbook] in the initial phases of the incident.” The agency emphasized the necessity of creating playbooks for “all expected requirements” to guarantee that organizations are prepared to react to a security breach instead of hurriedly crafting one in real-time.

The agency did not specify the extent of the delay in CISA’s response due to the absent playbook, and a spokesperson did not promptly reply to TechCrunch’s inquiry for comments. 

Independent cybersecurity reporter Brian Krebs mentioned in May that a security researcher from cyber firm GitGuardian alerted him to a multitude of exposed passwords located in a publicly available GitHub repository that had been uploaded by an employee of a CISA contractor.

As per Krebs, the researcher attempted to inform the contractor but received no response. Only after Krebs reached out to CISA did the agency take down the repository and revoke and replace all the compromised credentials to avert any potential misuse in the future.

CISA asserted that no customer or mission-related data was compromised in the incident and expressed gratitude to the researcher and journalist for their assistance. The agency admitted that its protocols for enabling security researchers to notify CISA of possible incidents “were not clearly defined,” and it has implemented changes to facilitate quicker communication for researchers wanting to contact the agency.

CISA has been without a permanent director since the commencement of President Donald Trump’s second term in January 2025. The agency has also experienced cutbacks, furloughs, and layoffs impacting around one-third of its workforce since Trump assumed office. 

Meta has discontinued a contentious feature that enabled users to alter images from public Instagram profiles using AI. The feature, introduced earlier this week alongside a suite of other AI functionalities, “didn’t hit the target” and is now off the table, as stated by the company.

This week, Meta unveiled Muse Image, a fresh AI image generator developed by Meta Superintelligence Labs, its specialized AI division. Meta highlighted one capability that allowed users to create images by @-mentioning public Instagram accounts they wished to cite. This feature, which wasn’t crafted to notify users if their pictures were utilized in this manner, sparked immediate criticism.

TechCrunch produced its own tutorial on how to turn off the feature.

Now Meta has changed its stance. The company released a blog post on Friday indicating it was retracting the feature. Puck News founding partner Dylan Byers was the first to report on the company’s decision.

“Our goal was to offer a beneficial creative resource and to empower individuals to decide if their public content could be referenced in this fashion,” the company stated on its blog. “We’ve received the feedback that this feature missed the mark, so it’s no longer available.”

TechCrunch contacted Meta for additional details and will refresh this article if they provide a response.

Since its adoption by social media networks, AI has often been misused — particularly to create explicit images of female celebrities. Platforms have tried to address this issue, although the measures taken have frequently been inadequate.

Concerning Meta’s recently removed feature, it appears quite clear that it could have been exploited. Indeed, Byers points out that the choice to eliminate the feature was made “in light of concerns from users and talent agencies, including CAA.”