Generative AI: Tech Startup Savior or Buzzy Nothingburger?
Spencer Rascoff serves as executive chairman of dot.LA. He is an entrepreneur and company leader who co-founded Zillow, Hotwire, dot.LA, Pacaso and Supernova, and who served as Zillow's CEO for a decade. During Spencer's time as CEO, Zillow won dozens of "best places to work" awards as it grew to over 4,500 employees, $3 billion in revenue, and $10 billion in market capitalization. Prior to Zillow, Spencer co-founded and was VP Corporate Development of Hotwire, which was sold to Expedia for $685 million in 2003. Through his startup studio and venture capital firm, 75 & Sunny, Spencer is an active angel investor in over 100 companies and is incubating several more.

Tech has been having a tough time. Widespread layoffs, a down economy, and a 2022 tech stock decline of 30% year-over-year all indicate the same thing: tech startups are in a bit of a slump. And yet generative AI is generating enough hype to put the VC world in a swirl, spurring hopes that ChatGPT, Dall-E and the rest can swoop in and resuscitate. Itās being touted as the next big thingāand it will be bigābut not in the way many people may hope or believe.
In the last few weeks Iāve been pitched dozens of startups that are trying to wedge their way into the spaceāeverything from AI-written real estate listings to customer service chatbots and beyond. Theyāve all been good ideas that could go on to help people and likely turn a profit. But I havenāt invested in a single one of them. Hereās why.
These startups are features, not companies. (I have written before about this startup trap.) Theyāre useful tools, some of which will likely become a part of our everyday lives, but from what Iāve seen, theyāre simply not robust enough to stand alone and definitely not addressing a large enough problem to achieve a venture-scale return.
A lot of the startup ideas in artificial intelligence now are simply capitalizing on generative AI functionality, being built as a thin, vertical layer of user interface that sits on top of large language models (LLMs). Even if the concepts coming my way have merit, I suspect theyāll be better incorporated into existing software and tech providers, rather than being built as an independent company. So despite all the buzz and media fervor, this isnāt going to be a gold rush for AI startups. Itās more like gold-leaf wrapped around an exciting tech development.
TripActions (recently rebranded as Navan) is a great example of this issue. Their core product is an intuitive travel expense management program and theyāre very successful at what they do. Itās a great concept that provides real value, and Iām sure there are currently dozens of startups trying to launch generative AI-focused expense management programs. But while theyāre clambering to get started, Navan is simply expanding to bring AI into their existing product. If thereās a legacy company like Navan thatās savvy and responsive, thatās going to be a big challenge for any disruptor trying to ride the AI wave to break into the market.
Itās also important to remember that this AI revolution did not start in November 2022 when chatGPT launched. While that was a seminal moment for layperson awareness, many companies have been using AI for years, and we have been unknowingly interacting with AI products. Spotifyās and Netflixās recommendations, Zillowās zestimate algorithm, and your smartphoneās camera āautocorrectā feature all use artificial intelligence to deliver their magic to you.
Iāve heard a lot of comparisons between generative AI and crypto or web3, but thereās actually a big gap between the two from a funding perspective. Crypto is great tech, but a lot of crypto-based startups werenāt fundable because there simply wasnāt a use case for their innovation. Generative AI startups might have the inverse problem. There are infinite use cases, myriad ways that it could automate and improve your work and creativity. But because most startup ideas for AI are essentially a veneer on top of core LLM tech, most of them arenāt ideas that are venture fundable.
Generative AI is much more akin to when the world pivoted to cloud. Like the shift to cloud computing, generative AI will impact everything. Itās foundational tech that operates in the background to make other things possibleāa staggering 94% of enterprises currently run on cloud services. I have no doubt that 10 years from now weāll say that without generative AI, there wouldnāt be X, Y, and Z. And just like cloud, which has a few big providers with huge data sets and huge revenue, I think there will be a few generative AI giants, including the LLM providers themselves. The rest of the ecosystem will sit on top of these models. Again: mostly features, not companies. Investors beware.
- 'Throw Everything Overboard That You Donāt Really Need': Spencer Rascoff Offers Founders Advice During a Downturn āŗ
- AI Chatbots Arenāt 'Alive.' Why Is the Media Trying to Convince Us Otherwise? āŗ
- At UCLA, Professors See 'Exciting Opportunities' in AI Writing Tools āŗ
Spencer Rascoff serves as executive chairman of dot.LA. He is an entrepreneur and company leader who co-founded Zillow, Hotwire, dot.LA, Pacaso and Supernova, and who served as Zillow's CEO for a decade. During Spencer's time as CEO, Zillow won dozens of "best places to work" awards as it grew to over 4,500 employees, $3 billion in revenue, and $10 billion in market capitalization. Prior to Zillow, Spencer co-founded and was VP Corporate Development of Hotwire, which was sold to Expedia for $685 million in 2003. Through his startup studio and venture capital firm, 75 & Sunny, Spencer is an active angel investor in over 100 companies and is incubating several more.