We’re experiencing a tale of two cities in the venture investing world for AI and non-AI companies. It’s been the best of times for pure-play AI startups and those that could convince investors they were close enough to being one. It’s been the worst of times for non-AI startups for the past two years; funding for some groups of founders has significantly plummeted and companies that would once have received follow-on venture rounds were left to become profitable or die.
With almost a quarter of all venture funding now flowing toward AI startups, I see the signs of some of investing’s darker days: a thin understanding of the core technology, the massive cash burn of the startups, poorly defined business models and sky-high valuations.
The signs, in other words, of a bubble.
For non-AI tech startups, almost the diametric opposite: The venture drought means investors now demand profitability or clear progress toward it or no additional funding is forthcoming.
If that seems vaguely schizophrenic, it is. Which of those investment strategies is right? And should founders, who rely on the expectation of future funding when starting companies, think twice before taking the startup plunge right now?
My answer: The large language model (LLM) train has left the station, but investors and entrepreneurs focused on startup fundamentals have a remarkable opportunity to create value. By staying focused on market opportunity, their business operations, the cost to acquire and retain customers, their cash burn, and — fingers crossed — their exit strategy.
Venture capitalists (VC) not running multi-billion dollar funds would be well advised to avoid new core large language model startups. But as for founders — there’s rarely been a better opportunity for startups, though it’s unlikely to be built on easy, ever-inflating rounds of venture funding.
The AI phenomenon has many of the earmarks of the dot-com and then the unicorn bubbles. The dot-com bubble also promised wild new world-changing tech, just like AI. And in the unicorn bubble, scale also matters greatly in AI. That means the biggest plays in AI are largely the province of big tech partnerships with giant VC firms. Microsoft, Amazon, Meta/Facebook and Alphabet/Google all are fighting proxy wars for AI dominance, largely through startups they’re funding in conjunction with the biggest names in the venture business like Sequoia, a16z, Tiger Global and others.
But as the LLM side of the AI bubble inflates, venture capitalists collectively should avoid slipping into some bad habits of the past. The tale of two cities is rearing its head toward certain classes of founders, leaving them to a lower status and success rate.
In 2023, all-female-founded companies raised $3.2 billion from VCs, a single-digit percentage of all U.S. VC activity. In comparison, all-male-founded companies raised $114 billion in 2023, according to a report from Pitchbook Data and Deloitte. Likewise, there has been a disturbing 71 percent drop in funding to diverse founders. Venture funding to Black-founded U.S. startups last year totaled only $705 million — marking the first time since 2016 that the figure failed even to reach $1 billion, Crunchbase data shows.
By not investing appropriately in Black women, we lost out on massive potential returns to the U.S. economy — returns that could have been used to improve public education, fund our first responders and military personnel, repair roads and bridges and more.
Are we destined to repeat and perpetuate that?
On AI’s other track, the discipline developed by the venture industry over the last two years meets the opportunities being unleashed by AI itself. OpenAI founder Sam Altman recently said he believed there will be an AI-powered billion-dollar valuation unicorn company with just one employee founded somewhere sometime soon.
It’s the incredible leveraging power of AI that could make that possible. Just as the two previous bubbles lowered the barriers to entry for startups through the proliferation of the Internet, cloud computing and mobile applications, AI brings the possibility of massive impact starting from a very small scale.
On that front, smart venture investing will be about sticking to the basics. Sourcing potential investments from places and through channels that others overlook. Buying in at favorable valuations and on favorable terms. Bringing smart due diligence to bear on business models, unit economics, and pathways to scale, and also making sure that management teams are strong and have the ability to get stronger.
Being able to piggyback on the infrastructure of large language models means that great ideas can truly come from anywhere. Entrepreneurs can bring brilliant insight or the knowledge of a deep customer need to the table and link to the extraordinary capabilities of AI to solve problems while building proprietary datasets that create more value and drive more optimization.
Investors who source from a richer, wider variety of origins can quickly realize the power of diverse founders and teams. There is gold in looking in different places than just Silicon Valley and Austin.
Those founders really are out there. (Our fund, for instance, recently backed an AI patent holder who is a Black woman from Virginia rather than the Valley. Her company is working on the next frontier for AI — Large Vision Models, which do for visual information what LLMs do for the written word.)
That’s technology that will benefit major segments of the population in areas such as health, smart homes, digital health and the aging population.
Perhaps diverse founders and management teams are more grounded and attuned to the realities of growing a business than the standard group of founders from the standard set of geographies and educational institutions. Why? Entrepreneurial success is about resilience, overcoming hardships, breaking the mold. Those are experiences women and minority founders know about firsthand.
But meanwhile — are we on the verge of an AI bubble? That depends. For those not writing $10 billion checks, it’s a different game. AI creates opportunities for outsized returns for companies leveraging the big LLMs — and smart venture investors will back founders from a wide range of backgrounds and geographies to help build innovative AI-powered companies.
If we want to see a thriving economy with profits and wealth generation, outsized returns for investors, exciting new technologies, products and services coming to market, along with job creation, renewed consumer confidence, rising household incomes and shrinking government deficits, the VC community needs to remain disciplined across the board about profitability, growth and performance.
It also needs to ensure it seeks opportunities with founders of all descriptions, not just those that fit traditional stereotypes.
If it does, the venture community will help ensure the best of times for founders, funders, customers and stakeholders of every description.
Gayle Jennings is the founder & CEO of Wocstar Capital, an early-stage investment fund focused on investing in tech companies with diverse and under-represented management teams.