The Trillion Dollar AI Question

When ChatGPT was released to the public in November 2022, it felt as though we had entered a new technological era, one that could change daily life in ways not seen since the internet. Disruption at that scale is exciting, particularly when it promises financial upside, and money quickly poured in. The clearest example was the Nvidia bull run, which delivered a stock return of returning nearly 700%.

Less visible to the general public, but equally significant, was the surge of investment across startups and venture capital. So far this year, AI-focused funding reached over $120 billion after already raising over $100 billion in 2024, and $55.6 billion in 2023. This chart by CB Insights does a great job at breaking down these numbers and showing how funding has been surging. The investment thesis is straightforward: AI tools will make companies and workers more efficient, lifting profits.

In addition to funding amounts, valuation premiums for AI companies are far surpassing those of solid B2B and SaaS companies, with the top 1% being valued at 3x-10x normal multiples. This graph from SaaStr does a really good job showcasing that current spread.

However, some interesting data published this past month from the think tank Model Evaluation and Threat Research has called some of this disruption ability into question. In a pre-experiment survey, experts predicted AI would speed developers’ work by nearly 40 percent. Afterward, participants thought it had made them about 20 percent faster. But when researchers measured actual output, developers completed tasks 20 percent slower with AI than without it. The study attributed this slowdown to the capability reliability gap, which is the gap between what models can theoretically do and how consistently they deliver useful results in practice.

In addition to this study, is the fundamentals on many of these AI companies that are getting huge cash infusions and massive valuations. While the top players, OpenAI and Anthropic, both are generating billions in revenue and growing very fast, profitability and cash burn is a major hair. It was reported this past weekend that OpenAI sharply raised its projected cash burn through 2029 to $115 billion as it ramps up its model training. For most AI startups, high burn has become the norm rather than the exception.

I don’t want you to come away from this thinking that I am just an AI hater. On the contrary, I do believe, in the same way that the internet did, that AI will ultimately fundamentally change the way we do a lot of things. Investors poured money in the 90s into basically any company with a “.com” name, and many people lost a bunch of money doing this.

One could argue that this go around the stakes are higher. In the first half of 2025, business spending on AI added more to GDP growth than all consumer spending combined. Many economists believe that the U.S.’s ability to weather the current economic headwinds has been a result of this AI spending alone.

Why This Matters

This likely won’t be the last time fears arise about a bubble before anything actually happens. Any time a sector starts getting a frothy there are multitudes of naysayers flying warning flags (crypto has been the poster child for this, and we will look more into this sector next month as there has also been a pretty interesting paradigm going on there). While the ultimate impacts of AI as a whole likely will be robust and disruptive as models continue to improve and adoption rates go up, the ultimate amount of disruption remains to be seen. Sam Altman, OpenAI CEO, I think summed it up best with this recent quote: "Someone is going to lose a phenomenal amount of money… smart people get overexcited about a kernel of truth.”

A measure of skepticism, I think, is always healthy

Tips of the trade

A section where we provide helpful tips for anyone with stock options or shares at private companies.

Funding your option exercise can be expensive and a require a large capital outlay. Feel free to reach out to us to discuss your options for partnering with ESO to exercise your options risk-free.

The ESO Fund does not provide legal, financial, or tax advice.

August's Top Ten:

1. Stubhub has been eyeing to hit the public markets for a while now, but misses in revenue continue to plague the company. The ticketing company disclosed in a securities filing that revenue grew about 3% to $828 million in the first six months of the year. That figure was 6% less than the $885 million it had anticipated for the first half earlier in the year.

2. OpenEvidence, an AI startup offering a ChatGPT-like tool for doctors to synthesize medical research, is looking to already double its July 2025 valuation to $6 billion. This escalation underscores the intense investor appetite for AI tools that promise to revolutionize health-care delivery, even as broader market volatility affects tech investments.

3. OpenAI boosted the size of its secondary share sale to a whopping $10 billion worth of stock. The transaction is being done at a $500 billion valuation.

4. Anthropic raised a $13B Series F this past month at a $183B valuation. Iconiq co-led the round with Fidelity and Lightspeed.

5. Databricks has confirmed that they raised a $1 billion round this past month at a $100 billion valuation. The round was led by Thrive, and comes as the company announced it reached $4B in ARR.

6. Vercel is in the process of raising at a $9 billion valuation, almost a 3x increase from their $3.25B valuation they achieved just over a year ago.  The round is reportedly being led by existing investor Accel.

7. Crunchbase data shows that total global venture funding in August plunged to $17 billion, the lowest monthly total since 2017. This represented a 12% year-over-year decline and a sharp 44% drop compared to July.

8. Lambda Labs has retained Morgan Stanley, J.P. Morgan, and Citi as it prepares for a U.S. IPO that could launch in the first half of 2026, riding strong GPU cloud demand.

9. Pinecone, the fast-growing vector database platform, is exploring a sale as mounting competition, especially after losing a major customer, signals shifting market dynamics.

10. Reflection AI, founded by ex-DeepMind researchers, is in talks to raise over $1 billion to develop open-source large language models that can rival efforts from Meta and others, with most of the capital reportedly already secured.

    Why It Matters:

    AI fundraising is showing no signs of slowing down, as we discussed in our intro. The jury is still out on how this AI market is ultimately going to shake out.

Startups that are hiring!

• Lovable

• Island.io

• Motion

• Chainguard

• Myrta

• Sierra

• Glean

• Abnormal

• Temporal

Open positions are per the company's website.

About ESO Fund

ESO Fund empowers startup employees to turn their stock options into reality. Since our inception in 2012, we've been dedicated to providing risk-free funding for the exercise of stock options, ensuring that individuals can seize the opportunities embedded in their equity.

Our mission is simple: to make equity compensation accessible and understandable. Through our innovative solutions, we've assisted countless individuals at 650+ companies in realizing the full value of their stock options, contributing to the success stories of numerous startup employees.

For more information on ESO Fund and how we can help fund your option exercise, please refer to our website at www.esofund.com!