AI is about to change search as we know it forever. But neither Google nor its competitors are getting it right.

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.

Bessemer Venture Partners ’ Talia Goldberg joined us for this week's VC Wednesdays. She shared what she looks for in early-stage startups, why 苹果 is the ‘dark horse’ in the AI race and shared advice for those just starting out in venture capital. Catch the full interview 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. Share your feedback and follow me for other tech updates.

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

A year ago, LinkedIn parent 微软 co-founder Bill Gates predicted that the rise of AI agents would one day make search engines redundant.

"Whoever wins the personal agent, that's the big thing, because you will never go to a search site again, you will never go to a productivity site, you'll never go to Amazon again," he said, speaking at a 高盛 and SV Angel event in San Francisco last May.

It seems as though we may very well be on the precipice of that massive shift.

Last week, reports surfaced that 谷歌 was mulling the idea of charging users for AI-based search — representing the biggest ever shake-up of its search business. Google is considering adding certain AI-powered search features to its premium subscription services, which already offer access to its new Gemini AI assistant in Gmail and Docs, the Financial Times first reported.

But why try and alter its $175 billion cash cow, which made up more than half of its total sales last year?

One obvious reason is that Google may be trying to recoup the massive cost of AI computing. AI models are known to be extremely expensive, requiring massive swaths of compute power to produce results.

But there is more than meets the eye, according to Shelly Palmer , professor of advanced media in residence at the S.I. Newhouse School of Public Communications at Syracuse University at Syracuse University .

“What is Google thinking?” he asked in an interview with LinkedIn News. “They're thinking that they don't want to be Kodak.”

'Innovator's dilemma'

While Google researchers pioneered the modern AI era with their pathbreaking paper on transformer models that underpin the technology behind large language models, the tech giant has been slow to respond to competitors like OpenAI gaining momentum since late 2022.

It has been testing an AI-powered search service called “Search Generative Experience” that provides more detailed answers to search queries in addition to presenting its signature links since last summer, but those experiments have yet to be added to its main search engine.?

Meanwhile, in the eyes of some experts, ChatGPT and competitive products like Perplexity ’s citation-backed search engine have been chipping away at Google’s dominance by giving users more complete answers in a matter of seconds — changing how people find information online, rendering Google’s traditional search lists of links and related ads potentially redundant and thereby, posing an existential threat to its ad-funded search engine business.

For the first time in the last two decades, cracks are starting to appear in Google’s search monopoly. Google’s search revenue was lower than Wall Street's expectations in its last earnings, and CEO Sundar Pichai avoided answering a question about whether that’s because fewer people are using Google Search, according to Stratechery.

Google today is essentially facing a classic case of innovator’s dilemma, tied between the necessity to move fast and innovate while figuring out how to maintain the customers and revenue stream from its money-making search engine business, said VC firm Bessemer Venture Partners ’ Talia Goldberg .

“Search is their golden goose, it’s virtually what fuels all of their profits and has made them one of the most successful and profitable companies in the history of the world,” she said in an interview with LinkedIn News. “And doing something that could potentially cannibalize that business — which is totally dependent on serving links, promoted posts and SEO-optimized content and replacing that with an instant answer — is really scary. That’s why they are thinking of charging for it, so as to not disrupt their golden goose.”?

It’s not like Google doesn’t have the means — the technology, the data and the talent — to figure out a novel way to monetize AI search, added Syracuse University’s Palmer. It’s just that despite these resources, it has botched the execution of many of its AI efforts so far. Just a month ago, Pichai wrote an email to staffers after its Gemini chatbot came under fire for producing historically and racially inaccurate images, saying "We got it wrong."

“Google has the power, the knowledge and the ability to be able to do this,” he said. “But they haven’t been able to package AI in any meaningful way.”

谷歌 , on its part, said that it was “not working on or considering” an ad-free search experience, but that it would “continue to build new premium capabilities and services to enhance our subscription offerings across Google.”

“For years, we’ve been reinventing Search to help people access information in the way that’s most natural to them,” a spokeswoman told LinkedIn News. “With our generative AI experiments in Search, we’ve already served billions of queries, and we’re seeing positive Search query growth in all of our major markets. We’re continuing to rapidly improve the product to serve new user needs.”

The challenges ahead for Google and all its competitors trying to rethink search

If Google decides to go ahead with charging for AI-powered search, it would not only have to contend with its enormous implications for search and the ad market around it as we know it, but would also have to take on the mammoth challenge of changing consumer behavior — convincing users to open their wallets for something that they have so far not had to pay for.

“Making this pivot is very difficult,” industry analyst Rob Enderle wrote in a recent report. “If Google doesn’t work to change perceptions and garner support, this effort is likely to fail spectacularly.”

And before it even starts attempting to change consumer hearts and minds, it will have to ensure that the product offers enough utility in the first place.

“Putting AI search behind a paywall seems premature. The user experience would have to be superb to meet expectations,” Nicholas McDonough , senior manager of SEO & strategy at?Verbatim Marketing Agency, wrote on LinkedIn. “The current UX with summaries is not appropriate in many cases. It would require a serious overhaul.”

Google is not alone in this quandary. Monetizing AI-powered chatbots has proved to be a more difficult task than selling ads against search queries for others as well.?

Perplexity , one of the fastest-growing consumer AI startups last valued at over $500 million, also seems to have recently learned that lesson. The startup boasts of having over 100 million monthly users including heavyweights like 英伟达 CEO Jensen Huang and Shopify CEO Tobias Lütke , but recently changed its tune and announced plans to sell ads, after initially touting that search should be "free from the influence of advertising-driven models” on its website.

The fundamental issue is that whether it’s incumbents like Google or startups like Perplexity, nobody has cracked the code on developing an interface suited to the age of AI quite yet. While a page full of blue links may not be the most suitable, a chatbot doesn’t fit the bill either. Investors like Bessemer’s Goldberg and Madrona ’s Karan Mehandru say that we’ll see new AI interfaces develop soon, especially as autonomous AI agents develop further.

“I doubt that a few years from now, prompting chat interfaces will be the dominant form of interacting with AI,” Goldberg said. “I'm excited to see companies innovating there, particularly a voice-first model or an ambient presence where AI is much more proactive in our lives, versus reactive.”

But who are they betting on? A tech giant that has thus far trailed even Google when it comes to delivering on AI — 苹果 — with whom Google is reportedly also in negotiations with, to build its Gemini AI models into its iPhones, in what is being dubbed a “blockbuster deal.”

“I'm pretty sure that Apple's working on a small LLM tailored and personalized to individual users — and they know more about us than any other company in the world because we have them sitting in our pockets all day,” said Mehandru.

Goldberg agreed, calling Apple the ‘dark horse’ in AI.

“They haven't done that much publicly yet, but they're so well-positioned, in part because they sit on an incredible data set,” she said. “But they also own the device and have their own chips, which is quite powerful, and can allow for much lower latency and higher performance.”

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

• 谷歌 , 英特尔 and Meta announced a slew of new AI chips this week, heating up the tech giants’ battle with Nvidia. Google will introduce a custom Arm-based server chip, called Axion, later this year, the company announced this week. Using an in-house chip — something other tech giants have already been doing for years — will help make cloud computing more affordable for Google, per CNBC. The business is a fast-growing source of revenue for parent Alphabet, though Google's share of the market still lags its rivals: It held 7.5% in 2022, while Amazon and Microsoft combined had about 62%. Google is also making more powerful versions of its artificial intelligence chips available to developers. Meanwhile, Intel is also debuting a new artificial intelligence chip. The company says that the new Gaudi 3 chip is 50% faster and "a fraction of the cost" of Nvidia's H100, and counts Dell, HP and Supermicro among the customers for Gaudi 3, which is set for a third-quarter release. On the other hand, Meta has rolled out a new generation of custom chips, aiming to ramp up its AI efforts and reduce reliance on outside suppliers. For now, the Meta Training and Inference Accelerator trains its ranking and recommendation algorithms, but the goal is to use MTIA for generative AI like its large language model, Llama.
• Speaking of chips, the push to expand U.S. chip manufacturing continues, with the government awarding up to $11.6 billion to 台积公司 . The $6.6 billion in grants and up to $5 billion in loans announced Monday will help the world's leading chipmaker, which supplies Apple and Nvidia, build three factories in Arizona. The funding includes $50 million to train local workers and will create more than 20,000 construction jobs, per Commerce Secretary Gina Raimondo. One of the new facilities will produce next-generation 2-nanometer chips, which Raimondo says are vital for AI and the military. TSMC will also increase its total U.S. investments from $40 billion to $65 billion as part of the deal. Intel was also recently awarded up to $20 billion in loans and grants for chip making facilities in four states, including Arizona; while Samsung too is expected to receive about $6 billion in grants.
• Jamie Dimon believes that AI will be as "transformational" for society as the printing press or electricity. The 摩根大通 CEO wrote in his annual shareholder letter that the technology could one day affect every one of the bank's 310,000 employees — whether by assisting them in their roles, or by making them redundant. Dimon's letter, released Monday, also contained a warning about inflation, predicting that U.S. interest rates could top 8% in the coming years amid "one of the most treacherous geopolitical eras since World War II." Meanwhile, 亚马逊 CEO Andy Jassy is vowing to keep costs down at the tech giant while also investing in AI. In his annual shareholder letter, published Thursday, Jassy emphasizes the importance of "primitive services" such as Amazon Web Services that are "foundational" systems on top of which new projects and businesses can grow. He says he believes that generative AI could be Amazon's next such "world-changing" service. Last month, Amazon upped its stake in AI startup Anthropic to $4 billion — the tech giant's largest outside investment ever. It has also scrapped a program that paid developers who created Alexa apps.
• OpenAI apparently doesn’t factor into Sam Altman’s wealth. Though the claim has been met with some skepticism, Sam Altman, one of 265 new billionaires whose membership in the “three-comma club” was recently made official by Forbes, did not make that first billion dollars the way many might assume. The OpenAI CEO has said he owns no equity in the ChatGPT maker, which is now worth more than $80 billion. According to Forbes, Altman’s swelling net worth can be attributed to early investments in startups including Reddit, which went public earlier this year, and Stripe, which was valued at $65 billion in a recent stock-sale.?
• Teachers are increasingly using AI to grade their students’ writing — a practice that offers many benefits, but also raises ethical questions. Half of college students used AI tools last fall, according to Tynton Partners, a strategy consultant firm. Among faculty members, usage more than doubled over the course of 2023, to 22%. Proponents argue AI can be like a teaching assistant, helping teachers grade faster and avoid getting burnt out; some critics say it's a potential breach of students' intellectual property, and should require their consent.

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

• Meta is planning to launch two small versions of its forthcoming Llama 3 large language model next week, according to The Information . The models come ahead of the launch of the biggest version of Llama 3, expected this summer.
• 谷歌 introduced a slew of products for its Cloud business customers at its Google Cloud Next event this week, including Google Vids, a new AI-powered video editing app that will become part of the company's paid productivity suite, Workspace, which will be released to Google's experimental Workspace Labs in beta in June. The company has also opened up Gemini 1.5 Pro for developers to build on top of. Meanwhile, OpenAI has also released an improved version of GPT-4 Turbo with Vision for developers to use.
• 甲骨文 and Palantir Technologies have announced a partnership to provide secure cloud and AI solutions aiming to help power businesses and governments around the world. The companies say that Oracle’s distributed cloud and AI infrastructure, combined with Palantir’s leading AI and decision acceleration platforms, will help users maximize the value of their data — contributing to increasing efficiency, addressing sovereignty requirements and helping them outpace adversaries.
• IBM has signed an AI collaboration agreement with the Government of Spain, IBM SVP and director of research Dario Gil shared on LinkedIn. The vision is to create the world's leading native Spanish LLMs, leveraging IBM's full-stack generative AI platform, which includes software, tools, and AI infrastructure, Gil said.
• AI model startup Symbolica AI launched this week with a $33 million Series A round led by Khosla Ventures . The startup says it is creating a brand-new architecture that it claims improves on transformers, the current best-in-class method used to build AI models.
• AI startup Alethea , which has an AI-powered tool that monitors social media for malicious disinformation campaigns, just raised a $20 million Series B round led by Google's venture arm GV (Google Ventures) .
• Spotify is beta-testing AI playlists, a feature that uses AI chat prompts to help you curate personalized playlists, CEO Daniel Ek shared on LinkedIn.

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

• TikTok is preparing to launch a new photo-sharing app to compete with Instagram , the company confirmed to TechCrunch . In recent days, TikTok users have received pop-up notifications that their existing photo posts will be shared on the new app, which will be called TikTok Notes. The news comes as TikTok faces a number of obstacles, including a potential ban in the U.S. Last year, TikTok lost its crown as the world's most-downloaded app to Instagram, which saw its downloads jump 20% over a year earlier. TikTok's rose 4% in the same period.?
• Tesla has settled a wrongful death lawsuit alleging that its Autopilot technology was responsible for the death of driver Walter Huang in 2018, according to court documents. Several lawsuits accusing Tesla of Autopilot failures are heading to trial this year, reports The Washington Post. The financial terms of the settlement were not made public. Autopilot is also under investigation by several federal agencies, including the National Highway Safety Administration, which has initiated more than 40 probes into the deaths of 23 drivers using the technology. The feature, which can help with steering and lane changes on highways, is available on all new Teslas. Speaking of Tesla, CEO Elon Musk is expected to announce a new Tesla factory and wider investment in India, Reuters reports.
• Internet service providers must now display information on the actual costs and speed of broadband services on their plans. These so-called nutrition labels, mandated by the Federal Communications Commission , will let consumers more easily comparison shop between plans, both online and at stores, and avoid hidden fees. Eight years in the making, the new rules had been firmly resisted by providers, but now Verizon , Google Fiber and T-Mobile have released their labels ahead of the April 10 deadline. Smaller ISPs have until Oct. 10 to comply. Consumer advocates are still calling on the FCC to break up the regional broadband monopolies that exist in rural or lower-income areas. Adding to pressure on low-income consumers is the looming expiration of a federal subsidy for internet service.
• Netflix 's new head of film is restructuring the division — an undertaking that's expected to include about 15 layoffs, Variety reports. Dan Lin , who recently took over from longtime chief Scott Stuber , is organizing the division by genre under four executives. Previously, Netflix film projects were split up by budget. After struggling "to manage Hollywood's biggest slate," the company is focusing on midsize comedies, rom-coms and family films, which are popular among streaming audiences, The Hollywood Reporter said. It is also planning "a handful of big movies and requisite awards contenders."?
• 苹果 is expanding in the Miami area, taking 45,000 square feet in a new building south of the city, Bloomberg reports, citing anonymous sources. That's in addition to a small office nearby that focuses mainly on the company's Latin American business, and a planned retail store in the $4 billion Worldcenter development. Apple follows other tech giants, including Amazon and LinkedIn parent Microsoft, in heading to South Florida since the pandemic. The surge in interest has pushed up Miami-area commercial property rents to almost $57 a square foot, per one commercial real estate adviser. Meanwhile, in India, Apple doubled its iPhone production last fiscal year to about $14 billion, Bloomberg reports, citing anonymous sources. The ramp-up is considered a sign that Apple wants to diversify beyond China as tensions mount between that country and the U.S.

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.

• Radha Iyengar Plumb has been appointed the new DoD chief digital and artificial intelligence officer.
• 纽约时报 ’ John-Michael Murphy has joined Perplexity ’s technical team, he announced in a post on LinkedIn.
• Ex-Meta news executive Campbell Brown has joined TollBit , a marketplace that connects AI bots and scrapers to publishers' verified content, Axios reports. Meanwhile, Meta has poached Snap Inc. AR exec Carolina Arguelles Navas to work on its AR expansion efforts, she shared in a post on LinkedIn.
• Operational Data Warehouse startup Materialize has hired Cockroach Labs ’ Nate Stewart as CEO, co-founder Arjun Narayan announced on LinkedIn. Narayan will take on the role of the startup’s VP of Engineering.?
• 谷歌 ’s VP of Android & Business Communications, Sanaz Ahari Lemelson and senior director of software engineering, Mike Curtis both announced that they are leaving the company on LinkedIn.?
• Spotify ’s global head of songwriter and publisher partnerships, Jules Parker , is leaving the company after eight years, he announced in a LinkedIn post.

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.