Cohere urges enterprise AI adoption, Rabbit says demand for its AI gadget is picking up, and more highlights from Collision this week

Welcome back to LinkedIn News Tech Stack, which brings you news, insights and trends involving the founders, investors and companies on the cutting edge of technology, by Tech Editor Tanya Dua. You can check out our previous editions here.

Pitch me the interesting investors, founders, ideas and companies powering emerging technologies like AI to reach the inboxes of nearly 1 million subscribers plus thousands more on LinkedIn. Follow me for other tech updates. And click 'Subscribe' to be notified of future editions.

A deep dive into one big theme or news story every week.

Nearly 40,000 attendees, 1,500 startups and hundreds of other tech companies gathered in Toronto for Collision for the last time this week, before it shifts base to Vancouver next year. 

The four-day conference was attended by tech veterans and disrupters alike — from Google Research’s vice president of strategy and operations Maya Kulycky to Cohere’s co-founder and CEO Aidan Gomez. 

I have been on the ground at the conference this week, meeting with entrepreneurs, investors and moderating a couple of panels. Here’s a rundown of the highlights so far:

Cohere says enterprises are charging ahead on generative AI

If 2023 was the year that businesses discovered generative AI, 2024 should theoretically be the year that they truly begin to deploy — and start seeing business value — from the technology. But as of earlier this year, 40% of enterprises still remained stuck in the exploration and experimentation phases of their projects, according to an IBM survey.

That is starting to change, said Aidan Gomez, the co-founder and CEO of Canadian OpenAI competitor Cohere, who spoke on stage on Tuesday.

“Last year, 2023, was the year of the proof-of-concept,” said Gomez, with companies launching a variety of pilots to experiment with the technology. “You’re starting to see this take off.”

Gomez said he was encouraged by early signs like the widespread adoption of AI co-pilots, which point toward businesses gearing up to buy the technology at scale. He also said that Cohere is now spending more on the compute needed to actually run the models for clients rather than merely training them. 

Cohere has a lot riding on broader enterprise adoption — and it’s not hard to see why. 

Unlike its competitors, the company does not have a consumer-facing chatbot and has always tried to appeal to enterprises, pitching itself as the right partner for large companies looking to privately deploy its models at scale for their individual needs. Plus, it just raised fresh funding of $450 million at a $5-billion valuation earlier this month. So, like other LLM-makers, it must demonstrate growth in revenue, not just to keep up with high compute costs, but also justify its lofty valuation.

But Gomez shouldn’t hold his breath; he may have to wait a little longer for the technology to truly take off in the enterprise. Elsewhere this week, McDonald's shelved the AI drive-thrus it had been testing at more than 100 locations in partnership with IBM, after videos surfaced showing customers struggling to order with them.

AI hardware maker Rabbit says demand is picking up

While large language model makers and other AI software startups have seen an influx of funding and fanfare in recent years, the same has not necessarily been true for AI hardware startups, which have struggled to gain traction.

Jesse Lyu, founder and CEO of the AI device maker rabbit inc., defended the startup on Tuesday following the launch of its R1 device earlier this year, saying that most initial reviews were based on an early version of the product, which the company has iterated on and updated since.

“It’s been a rough early release, there’s no doubt about it — but as a startup, this is the best result we can get,” he acknowledged. “Most of the early reviewers’ concerns have already been addressed,” he continued, adding that the company was fixing bugs and launching updates on a weekly basis.

Lyu also said that it was unfair for startups like his to be held to the same standards as bigger tech giants’ hardware products. He added that Rabbit has now sold 130,000 of its R1 devices, with usage doubling every two weeks. Rabbit has raised $30 million so far from backers including Khosla Ventures.

“We were expecting to sell 10,000 units — we’re now at 130,000 units worldwide,” he said.

Priced at $199, the R1 is a handheld AI gadget equipped with a tiny touchscreen, a mic as well as a camera that connects to cellular networks or wifi, and lets users not just ask questions but can also take actions on their behalf. Lyu, for example, demonstrated how the device can be used by people to quickly figure out whether you can park in San Francisco at a given time or not, which can vary depending on the day and time.

Billionaire businessman Frank McCourt backs warning labels for social media

In a New York Times op-ed published Monday, U.S. Surgeon General Vivek Murthy called for warning labels for social media to alert teens and their parents to the mental-health risks associated with its use. 

Frank McCourt, the founder of McCourt Global and former owner of the Los Angeles Dodgers who is vying to buy TikTok in the U.S., took things a step further, calling social media “an epidemic” at Collision this week.

“This is an epidemic, and our responsibility as adults here is to protect the next generation,” he said. “I am really, really concerned — borderline petrified — about what my younger kids are entering into right now.”

But while he thinks that social media is fundamentally harmful, he did not back government oversight of these platforms.

“The policy-making, regulatory environment cannot possibly keep up with the speed and the power of technology,” McCourt said.

Instead, he touted his plan “Project Liberty” as the way forward, which seeks to rebuild the web in a decentralized manner using blockchain technology, giving users direct control over their data and digital identities. That principle is also why his bid to purchase TikTok excludes the platform's powerful recommendation algorithm.

“We don't want it and we don't need it, which is why our bid has a great chance of succeeding,” he said. “We don't think replacing China's control and money with some other sovereign's control and money solves anything.”

McCourt, of course, is not the only one who has his sights set on TikTok’s U.S. operations, if the company were to offload its business to comply with a U.S. law that’s threatening to ban it. Former Treasury Secretary Steven Mnuchin and former Activision CEO Bobby Kotick, respectively, have also floated bids to purchase the company.

Here’s where we bring you up to speed with the latest advancements from the world of AI.

• OpenAI co-founder launches new venture. Ilya Sutskever, a co-founder of OpenAI who departed last month, has announced his new project: developing an AI system with an emphasis on research and safety. Sutskever's new firm, called Safe Superintelligence, has no near-term plans to sell AI products or services, he said, and won’t be in a “competitive rat race.” Sutskever, a central figure in OpenAI CEO Sam Altman’s ouster and subsequent return last year, declined to say how much the new company has raised or who its backers are.

• Meanwhile, OpenAI is reportedly eyeing a new structure. In a move that could pave the way for an IPO, OpenAI CEO Sam Altman is looking at changing its structure to a for-profit benefit corporation, The Information reports, citing anonymous sources. The new structure, in use by rivals xAI and Anthropic, would eliminate control currently held by OpenAI’s nonprofit board over its for-profit unit. A recent struggle that resulted in Sam Altman’s temporary ouster exposed the "fragility" of the current unusual structure, per the Information. The new one would let OpenAI maintain a mission — using AI to benefit humanity. Since late last year, OpenAI has more than doubled its annualized revenue to $3.4 billion, CEO Sam Altman has told staff.

• Google’s AI push is sparking tensions internally. As the race to bring AI products to market intensifies, Google’s efforts are increasingly under scrutiny following a series of high-profile missteps. After combining two of its AI units, DeepMind and Google Brain, over a year ago, friction is emerging as teams that have historically been focused on research pivot to commercialization strategies, Bloomberg reports, citing anonymous sources. Some AI workers at the company have felt “frustrated" about having to follow specific product road maps that potentially risk disrupting Google’s strength in foundational research. 

• Apple isn’t paying for its ChatGPT deal. In fact, the partnership between the tech giant and OpenAI isn't expected to generate "meaningful revenue" for either company, at least initially, according to Bloomberg's anonymous sources. Instead, Apple sees the value of pushing OpenAI’s brand and technology to its millions of users as "of equal or greater value than monetary payments," while Apple benefits from customers spending more time on devices or upgrades, the sources said. That doesn't mean there couldn't be pay deals later, especially as the tie-up could become expensive for OpenAI in terms of cloud-computing.

• It's official: NVIDIA has become the world's most valuable public company. Its market cap climbed to $3.34 trillion on Tuesday, surpassing LinkedIn parent Microsoft and Apple, which had been jockeying for the top spots in recent weeks. Corporate and investor enthusiasm for generative AI has helped its stock surge over 170% so far this year, reflecting a gain of over $2 trillion. Nvidia is seen as major beneficiary of the AI boom, since it currently has about 80% of the market for the chips used in data centers. It surpassed Apple's market value earlier this month. But for some, Nvidia’s unrelenting rise is evoking memories of the dot-com boom. 

Here’s a list of other notable AI developments from this week: 

• Toronto-based self-driving startup Waabi announced fresh funding of $200 million on Tuesday, and touted a 2025 rollout of the company’s generative AI-powered driverless trucks. The Series B funding round was led by Uber and Khosla Ventures, with new strategic investors including Nvidia, Volvo Group Venture Capital, Porsche Automobil Holding SE, Scania Invest and Ingka Investments. See CEO Raquel Urtasun’s LinkedIn post here. 

• Meanwhile, Amazon-backed AI startup Anthropic has released a new version of its Claude model called “Claude 3.5 Sonnet” that it says surpasses the capabilities of GPT-4o, OpenAI’s most recent AI system.

Catch up on the tech headlines you may have missed this week and what our members are saying about them on LinkedIn.

• Meta is restructuring its Reality Labs division. Meta's hardware division is undergoing its biggest restructuring in four years, The Verge reports, citing an internal memo. Reality Labs will now be organized into two groups, "Metaverse," which includes its Quest headset line, and "Wearables," including Meta's recent surprise hit, its Ray-Ban smart glasses. In the memo, CTO Andrew Bosworth declared, "[W]e are doubling down on finding strong product market fit for wearable Meta AI, building a business around it, and expanding the audience." Some employees have reportedly been laid off in the reorganization, but it's not clear exactly how many.

• Meanwhile, Apple has halted work on its high-end Vision Pro headsets, amid weak sales of the $3,500 product, The Information reports, citing an anonymous source. The company is still aiming to release a more affordable Vision model by the end of 2025, according to the report, and originally planned to market two versions of the headset — similar to the regular and Pro iPhone models. Apple's shift in focus could influence competitors such as Meta, which see the Vision Pro "as a litmus test of consumer appetite for a premium device." Apple is also shutting down its Pay Later program about a year after launching the service. Instead, third-party lenders, including Citigroup and Affirm, will be integrated into the upcoming iOS 18 software release.

• The U.S. has sued Adobe over subscriptions. The Federal Trade Commission says that the software giant made it too hard for customers to cancel their subscriptions, effectively trapping customers “into yearlong subscriptions through hidden early termination fees and numerous cancellation hurdles.” The regulator also claims Adobe’s tactics, such as “hiding” the terms of its plan in the “fine print and behind optional textboxes and hyperlinks," violate a 2010 consumer protection law. The U.S. government started looking at Adobe’s cancellation policies last year. 

• Elon Musk tries to mollify X advertisers at Cannes Lions. At The New York Times’ Dealbook Summit last year, the X owner directed an expletive at advertisers from the stage. At the Cannes Lions festival this week, he attempted to make nice. In front of an audience of marketing professionals, Musk said his comment was about free speech and he understood their desire to "appear next to content that they find compatible with their brands." At the time, major advertisers were fleeing X after Musk expressed support for an antisemitic post. He highlighted X's advances in AI-driven ad targeting and its "influential" user base, urging advertisers to reconsider. National ad spending on X is down 65% since May 2023, compared with the year before, while every other social media outlet gained, The Information reports, citing data from Guideline. X CEO Linda Yaccarino is also at Cannes trying to woo advertisers, touting initiatives such as sports docuseries.

Here’s keeping tabs on key executives on the move and other big pivots in the tech industry. Please send me personnel moves within emerging tech.

• OpenAI has named former NSA head and retired U.S. Army Gen. Paul Nakasone to its board and its safety and security committee, as it looks to ensure its models are safe as it works toward its goal of super intelligence.

• AI startup Writer has hired former Coupa executive Roger Kopfmann as its first CFO, Andy Shorkey as its chief revenue officer and Diego Lomanto as its chief marketing officer. See Writer CEO May Habib’s LinkedIn post here.

Thanks for reading. Please share Tech Stack and forward it around if you like it! And if you have any news tips, find me on InMail.