Techin' Geek

Three unidentified plaintiffs contended in a lawsuit submitted on Monday in California federal court that Elon Musk’s company xAI must be held liable for permitting its AI systems to create abusive sexual images of recognizable minors.

The trio of plaintiffs seeks to initiate a class action lawsuit on behalf of any individual whose genuine images as minors have been transformed into sexual content by Grok. They claim that xAI failed to implement fundamental safety measures employed by other leading laboratories to stop their image models from generating pornography involving real individuals and minors.

The legal case, Jane Doe 1, Jane Doe 2, a minor, and Jane Doe 3, a minor versus x.AI Corp. and x.AI LLC, has been filed in the United States District Court for the Northern District of California.

Other deep learning image-generating technologies utilize diverse strategies to avert the production of child pornography from ordinary photographs. The lawsuit asserts that xAI did not embrace these standards.

Significantly, if a model is capable of generating nude or erotic content from real pictures, it becomes nearly impossible to prevent it from creating sexual content featuring children. Musk’s public endorsement of Grok’s capability to produce sexual imagery and portray real individuals in revealing outfits is a fundamental aspect of the case.

The company has not commented on a request from TechCrunch.

One plaintiff, Jane Doe 1, had images from her high school homecoming and yearbook altered by Grok to show her naked. An anonymous informant who reached out to her on Instagram alerted her that the images were spread online, providing her with a link to a Discord server containing sexualized pictures of her and other minors she recognized from school.

A second plaintiff, Jane Doe 2, was alerted by law enforcement about altered, sexualized depictions of her generated by a third-party mobile application that utilizes Grok models. A third individual, Jane Doe 3, was also informed by law enforcement who found an altered, pornographic image of her on the device of a suspect they had detained. The plaintiffs’ attorneys assert that since third-party utilization still necessitates xAI code and servers, the company should be deemed responsible.

All three plaintiffs, two of whom are minors, report severe distress related to the dissemination of these images and the implications it may hold for their reputations and social interactions. They are requesting civil penalties under various laws designed to safeguard exploited children and avert corporate negligence.

During the Nvidia GTC keynote on Monday, CEO Jensen Huang unveiled DLSS 5, an advanced iteration of the chipmaker’s AI graphics technology aimed at enhancing video game realism while reducing computational requirements. 

The innovative DLSS 5 framework integrates conventional 3D graphics data with generative AI models capable of predicting and completing sections of images, enabling Nvidia’s GPUs to render intricate scenes and realistic characters without having to recreate every detail from scratch. 

“We merged controllable 3D graphics, the authentic representation of virtual environments, the organized data… with generative AI, probabilistic computing,” Huang stated in his keynote. “One is entirely predictive, while the other is probabilistic but highly realistic.”

Huang remarked that the fusion of these two elements—structured data and generative AI—empowers developers to craft content that is “stunning, incredible, and also manageable.”

“The integration of structured information with generative AI will recur across various sectors,” Huang predicted. “Structured data serves as the backbone of reliable AI.”

Although gaming currently constitutes a smaller segment of Nvidia’s revenue compared to its historical significance, it remains the industry that established Nvidia’s trajectory. Huang presented DLSS 5’s methodology as a representation of a wider computational transition, implying its potential application beyond gaming into enterprise computing. 

The billionaire leader highlighted enterprise data solutions like Snowflake, Databricks, and BigQuery as instances of structured datasets that upcoming AI systems may analyze and derive insights from. 

“Looking ahead, these data structures will be leveraged by AI, which will operate at a significantly higher speed than humans,” Huang declared. “Future agents will utilize both structured databases and unstructured, generative databases. These databases encompass the vast majority of global information.”

Apple’s latest MacBook Neo is not only the least expensive MacBook yet — it also ranks as the most repairable MacBook in “approximately fourteen years,” based on a thorough teardown by the how-to site iFixit.

According to iFixit, the “key highlight” is the battery. While previous MacBook batteries were allegedly glued in, the Neo’s battery is located in a tray fastened with 18 screws. Although this is a considerable number of screws, iFixit stated that “screws always outperform adhesive.” This new setup — which should facilitate battery replacement for users — “elicited cheers throughout the iFixit workspace.”

Additional modifications that supposedly enhance the repairability of the MacBook Neo include a straightforward disassembly tree, the Repair Assistant’s acceptance of replacement components without issues, and a display and keyboard that are simpler to replace.

In the end, iFixit had enough reservations — such as soldered RAM and storage — to assign the Neo a repairability rating of 6 out of 10. They noted that for a MacBook, “this is a solid score.”

[embedded content]

In 2012, Apple revealed a redesign of its MacBook line featuring the new Retina Display MacBook Pro. This redesign marked a shift towards handling repairs internally or through Apple-endorsed shops, sacrificing much of the DIY repairability offered by previous versions. Many components in the new MacBooks, such as memory cards and batteries, were glued down; iFixit rated that model a 1 out of 10 for repairability.

Since that time, Apple has been criticized for its opposition to certain right to repair legislation in various states and for creating what are referred to as parts-pairing barriers, which restricted or prevented devices from functioning with non-manufacturer components. Legislative changes that permitted individuals to repair gadgets in states like Oregon prompted the company to ease limitations on repairing iPhones with older components.

The new MacBook Neo still has soldered memory and storage, complicating upgrades. “However, the parts that typically fail first are more accessible than they have been in many years on a MacBook,” iFixit remarked. “The battery is secured with screws rather than glue. The ports are modular. The display is easier to replace. The internal layout is surprisingly logical.”

Nvidia is set to commence its yearly GTC developer conference in San Jose, California, on Monday, with CEO Jensen Huang’s keynote planned for 11 a.m. PT / 2 p.m. ET.

GTC — which signifies GPU Technology Conference — is Nvidia’s premier yearly gathering, taking place from March 16 to March 19. The semiconductor giant generally utilizes this platform to reveal new products, highlight collaborations, and present its vision for the future of computing. Huang’s keynote will emphasize Nvidia’s contribution to the future of computing and AI. Attendees can view the two-hour address live at the SAP Center or stream the presentation on the event’s website. The YouTube livestream is integrated below.

[embedded content]

The extensive four-day event concentrates on future AI developments across various sectors, such as healthcare, robotics, and self-driving vehicles.

On the software front, there are speculations that Nvidia may unveil an open-source platform for enterprise AI agents, referred to as NemoClaw, as initially reported by Wired. This platform would provide companies a systematic approach to create and implement AI agents (software capable of performing complex tasks independently) and would position Nvidia to align with similar offerings from firms like OpenAI.

On the hardware front, the company is also believed to be introducing a new chip intended to enhance the AI inference process — the method whereby an AI model utilizes learned information to generate answers or make choices, distinguishing it from the initial training phase that demands significantly more computational resources. Accelerating and reducing the cost of inference is widely recognized as one of the remaining hurdles to broad AI application scalability. This chip would signify Nvidia’s latest effort to lead not only the training sector, where it already holds an estimated 80% market share, but also the inference market, where custom chips from Google, Amazon, and others are rapidly emerging as competitors.

Moreover, a variety of partnership announcements and demonstrations highlighting Nvidia’s AI capabilities across sectors are expected.

Kevin Cook, a senior equity strategist at Zacks Investment Research, informed TechCrunch that attendees should also anticipate insights into what the company intends to do regarding its partnership with Groq, the inference firm Nvidia reportedly invested $20 billion in late last year to acquire its technology. There is considerable interest around this collaboration, particularly since Jonathan Ross, Groq’s founder; Sunny Madra, Groq’s president; and additional members of the Groq team have agreed to join Nvidia to further develop and scale that licensed technology.

Next week, Nvidia will launch its annual GTC developer conference in San Jose, California, featuring CEO Jensen Huang’s keynote on Monday at 11am PT / 2pm ET.

GTC, which represents GPU Technology Conference, serves as Nvidia’s primary yearly gathering, where the chip manufacturer often announces new products, promotes partnerships, and outlines its vision for the future of computing. Huang’s keynote is set to emphasize Nvidia’s contribution to the future of AI and computing. Attendees can view the two-hour speech live at the SAP Center or stream it online via the event’s website.

The comprehensive three-day event will concentrate on emerging developments in AI across various sectors, including healthcare, robotics, and self-driving vehicles, among others.

In the realm of software, there are speculations that Nvidia plans to unveil an open-source platform for enterprise AI agents, referred to as NemoClaw, as initially reported by Wired. This platform is expected to offer businesses a structured approach for creating and deploying AI agents (software capable of executing multi-step tasks independently) and would position Nvidia to align with similar products from companies like OpenAI.

Regarding hardware, the company is also said to be launching a new chip intended to enhance the AI inference process — the mechanism by which an AI model utilizes its acquired knowledge to produce responses or make decisions, differing from the initial training phase, which demands significantly more computational resources. Achieving faster, more cost-effective inference is generally viewed as one of the final hurdles to the widespread scalability of AI applications. This chip, if verified, would mark Nvidia’s latest effort to dominate not only the training sector, where it already holds an estimated 80% market share, but also the inference arena, which faces increasing competition from custom chips developed by Google, Amazon, and others.

Kevin Cook, a senior equity strategist at Zacks Investment Research, informed TechCrunch that participants should anticipate insights regarding the company’s plans concerning its partnership with Groq, the inference firm Nvidia reportedly acquired for $20 billion last year to license its technology. There is significant interest surrounding this collaboration, especially considering that Jonathan Ross, Groq’s founder, Sunny Madra, Groq’s President, and several members of the Groq team have agreed to join Nvidia to advance and scale the licensed technology.

Additionally, a variety of partnership announcements and demonstrations will highlight Nvidia’s AI capabilities across different industries.

Rox, a startup focused on creating autonomous AI agents to enhance sales productivity, has secured a new funding round that values the company at $1.2 billion, as reported by various sources.

This funding round featured a leading investment from returning supporter General Catalyst, two sources indicated. Rox and General Catalyst did not reply to TechCrunch’s inquiry for comments.

At the time of the fundraising, which concluded last year, Rox was anticipated to finish 2025 with an annual recurring revenue (ARR) of $8 million, based on insights from two individuals acquainted with the transaction.

In November 2024, Rox revealed it had secured a cumulative $50 million, comprising a seed round spearheaded by Sequoia and a Series A round led by General Catalyst, with contributions from GV.

Founded in 2024 by Ishan Mukherjee, the former chief growth officer of New Relic, Rox emerged after Mukherjee joined New Relic due to its acquisition of Pixie, a software monitoring startup he co-founded, in 2020.

The startup promotes itself as an intelligent revenue operating system that integrates with a company’s existing software systems—ranging from Salesforce to Zendesk—deploying numerous AI agents. These agents assess existing accounts, investigate potential clients, and refresh CRM software. By centralizing these functions, Rox strives to replace and simplify the multitude of disjointed software tools presently utilized by sales teams.

“Rox’s distinctive AI agent system enhances the CRM experience,” GV investor Dave Munichiello stated in a 2024 blog post when discussing the Series A funding. “These agents work tirelessly behind the scenes to track customer interactions, identify possible threats and opportunities, and even recommend the optimal course of action.”

Rox faces competition across various sectors, including well-established revenue intelligence firms like Gong and Clari, alongside AI sales development platforms such as 11x and Artisan. Additionally, a continuous influx of new AI-native, comprehensive CRM competitors is entering the market, such as Monaco—a startup initiated by Sam Blond, the former president of corporate spending platform Brex—which emerged from stealth mode last month.

Based on Rox’s website, the firm counts Ramp, MongoDB, and New Relic among its clientele.  

There’s an exciting new approach to enjoy YouTube: by channel surfing like a nostalgic viewer with cable TV. This inventive concept originates from London developer Steven Irby, who has recently unveiled a web app named Channel Surfer, which displays captivating YouTube videos in an interface that mimics a vintage TV guide.

Within the app, you can explore various topic-centric channels and click to tune in as if you were watching live television.

At its debut, there are 40 custom-made “channels” available, covering broad topics such as news, politics, sports, and lifestyle content, alongside a variety of music channels and those focused on technology.

The tech-oriented group features channels such as “AI & ML,” “Code & Dev,” “Space,” “Retro Tech,” “Tech & Gadgets,” and “Gaming.”

As you navigate through channels, you join the video currently playing in the middle of its stream. At the same time, the guide keeps you updated on upcoming content across all channels and the scheduled times for viewing. You can also scroll forward to see programming scheduled for the next 24 hours.

This creates a viewing experience similar to that of traditional live television — an approach that has gained popularity on free streaming platforms like Plex, Pluto TV, Tubi, and others, which feature live channels offering TV shows and movies. YouTube, for its part, remains the leader in TV streaming within the U.S.

Additionally, a small counter at the bottom of the screen shows how many other users are currently watching YouTube alongside you.

Irby mentions that he developed this concept to recreate a similar experience at the streamers, but specifically for YouTube videos, considering that discovering something appealing to watch can still be challenging.

“I created Channel Surfer because I’m weary of the algorithms and decision fatigue,” Irby remarked to TechCrunch. “I long for the days of channel surfing without the pressure of choosing what to watch. I want to relax and watch what’s on without contemplating what to select next.”

“My boomer Mom is a cable TV viewer. I desire the same experience but with my YouTube channels. Additionally, it’s oddly comforting to realize I’m watching alongside other viewers,” he expressed.

This initiative is part of Irby’s series of fresh experiments, as he is a 40-year-old tech industry expert who has spent the last decade traveling globally.

“I have an abundance of creativity arising from my long and eclectic journey. I can’t tolerate the notion of being a Jira ticket worker any longer,” he stated.

The app appears to be quite popular, with Irby noting that Channel Surfer’s newly launched website attracted over 10,000 views on its debut day.

Behind the scenes, Channel Surfer is currently a static Next.js site utilizing PartyKit and is hosted on Cloudflare. The channels and music it provides are selected by Irby himself. GitHub Actions runs a script that updates the data daily. There’s no existing back end yet.

And while Claude aided in the coding process, Irby emphasizes that the site is not “vibe-coded.”

The channels essentially feature YouTube embeds, inclusive of YouTube’s ads, so the app should comply with policy. Eventually, Irby hopes to extend the app to TV platforms like Fire TV, Google TV, and others. (It is also compatible with mobile devices and tablets, though it requires further development.)

At launch, Channel Surfer is a complimentary service that grants access to 175 YouTube channels and 25 music playlists. However, by subscribing to Irby’s newsletter, users can import their own YouTube subscriptions into the app.

The process is straightforward: You drag a “Channel Surfer” bookmarklet to your bookmarks bar, open your YouTube subscriptions, and click the bookmarklet. The process starts, directing you back to the app where you paste the copied JSON text into a designated box and hit the “import” button. This integrates your channels into Channel Surfer’s current lineup, potentially adding hundreds more channels to view in this manner.

The site’s existence reflects the web’s earlier days, brimming with creative experiments and fun. For Irby, that’s the intention.

“I’m passionate about demonstrating that the old web is still thriving,” he claims. “It’s simply hidden beneath a pile of clutter.”

On Thursday, Rivian unveiled the specifications and pricing information for its crucial R2 SUV, and the firm finally addressed a long-standing inquiry: When will buyers be able to purchase the anticipated $45,000 entry-level model?

The response is “late 2027,” as per the company’s press announcements. However, there’s a stipulation. The terminology Rivian is currently using indicates that the base model R2 will be “starting around $45,000.” This is a significant shift from the way the company had been previously marketing the R2 as “starting at $45,000” on its website. (Emphasis mine.)

This development is not exactly unexpected. As reported by TechCrunch last week, Rivian eliminated the “starting at $45,000” phrasing from its website in February.

Moreover, many factors have transformed since Rivian initially introduced the R2 in March 2024. The $7,500 federal EV tax credit is no longer available. Established automakers have ceased purchasing regulatory credits from companies like Rivian, effectively terminating a flow of what appeared to be free financial resources inflowing into its finances. Former President Trump’s disruptive tariffs have escalated the expenses of components and materials that Rivian utilizes for manufacturing its EVs.

In certain respects, Rivian faces even more significant obstacles to confront.

Sales of its R1T truck and R1S SUV experienced a downturn in 2025. Rivian is preparing to commence the construction of a massive factory in Georgia, where it anticipates producing hundreds of thousands of R2 SUVs (and eventually, R3 hatchbacks).

The firm is also attempting to orchestrate what could become one of the swiftest electric vehicle launches in American history with its more upscale R2 versions this year. Rivian is forecasting sales between 20,000 and 25,000 R2s by the conclusion of 2026. If successful, only Tesla’s Model Y would have achieved 20,000 in sales at a faster pace.

Rivian informed TechCrunch that it aimed to commence with the more expensive performance R2 models “so owners can experience the absolute peak of the new platform first.”

“Launching with a high-spec trim is a standard practice in the industry and establishes the stage for the entire lineup by demonstrating the remarkable capability and acceleration that make a Rivian unmistakable, all while we ramp up production into our Premium and Standard configurations afterward,” stated the company.

Rivian plans to release a “Standard” R2 in the first half of 2027 that will begin at $48,490, with a range reaching up to 345 miles. The actual base model will only achieve around 275 miles. This could indicate how Rivian is approaching the base model cost — fewer batteries typically correlate with reduced expenses. The base model’s more limited range may also serve a dual purpose by motivating customers to spend several thousand dollars more for noticeably better range.

Rivian indicated to TechCrunch that the two Standard models share the same rear-wheel-drive system but did not specify if there are other distinctions besides the battery capacity that would account for the price variation. It also opted not to comment on its upselling tactics.

“We have implemented substantial internal engineering, development, and business initiatives to achieve our target price. We streamlined complexity by transitioning to a zonal electrical architecture, minimizing the number of electronic control units, and employing our in-house drive units,” the company stated. Rivian noted it also utilized lessons learned from reducing costs in its second-generation R1 vehicles and improved supplier relationships.

This messaging appears just months after Rivian consented to a $250 million settlement regarding a class action shareholder lawsuit tied to the sudden price increases for its R1 vehicles back in 2021.

It also echoes some of the controversies Tesla found itself in a few years ago. Elon Musk and his firm had spent a significant amount of time promising that the Model 3 would retail for $35,000. However, Tesla offered a $35,000 Model 3 for a brief period “off-menu,” and that initiative did not endure long. Many customers who attempted to purchase it were pressured into opting for higher-trim versions, while Musk publicly lamented the challenges in meeting his prior promise.

Another Tesla model was initially announced at an appealing price that never materialized: the Cybertruck. Tesla originally promoted the steel-bodied pickup in 2019 as starting at just $40,000. Ultimately, it launched at much steeper prices that, combined with its widely unappealing design, resulted in disappointingly low sales.

It seems improbable that the R2 will face as severe a fate as the Cybertruck did for Tesla. After all, it’s a significantly more accessible vehicle that is also starting at a much lower cost — all without the political ramifications of having Elon Musk as the CEO. However, only the coming years will reveal whether the R2’s entry model will follow a trajectory more akin to the $35,000 Model 3, the Cybertruck, or something entirely distinct.

The publishing service Substack continues to enhance its focus on video content with the introduction of the Substack Recording Studio, a built-in feature that allows creators to pre-record and publish video content.

Exclusively available on desktop, the studio facilitates individual videos as well as discussions with up to two guests. Creators have the option to add personalized watermarks and can share their screen with co-hosts. After recording, Substack automatically generates clips and thumbnails for users to share.

“Previously, producing video on Substack required going live or using multiple tools: a recording application, a platform for creating and sharing clips, and a tool for designing thumbnails,” the company mentioned in a blog entry. “Substack Studio consolidates all these functions into one hub.”

The entry further highlights that creators who have leveraged audio or video on Substack in the last 90 days have seen revenue growth 50% quicker than those who have not.

While Substack is primarily recognized as a newsletter service, the company has exhibited a strong interest in video in recent years, focusing on developments that position it more as a competitor to Patreon, motivating creators to delve into multimedia.

Since permitting creators to upload videos in 2022, Substack expanded capabilities to include livestreaming and monetization of videos last year and subsequently introduced a $20 million Creator Accelerator Fund to assist creators in moving from alternative platforms to Substack.

Similar to Instagram, Substack has also recently rolled out a TV application, now available on Apple TV and Google TV. This app allows audiences to view video posts and livestreams on television and features a TikTok-style “For You” section, offering additional recommendations.

Even with the rising trend of consuming short-form videos on smartphones, audiences appear to be opting for TV screens to view longer content. Netflix has been heavily investing in the integration of video podcasts into television. On YouTube, viewers engaged with over 700 million hours of podcasts monthly on living room devices (including TVs) in 2025, marking an increase from 400 million hours monthly in the previous year.