OpenAI After the Mega Funding Round. Is the Mega IPO Coming?
ChatGPT's creator is now worth $157 billion, despite racking up billions in losses. Is it an IPO candidate or a bottomless pit?
Sam Altman, Co-Founder and CEO of OpenAI, has done it again. On 2 October, he announced the completion of the latest round of funding. He was able to raise 6.6 billion USD from investors at a gigantic valuation of 150 billion USD (pre-money). This means that the company behind ChatGPT is now valued at 157 billion USD, including the newly raised cash.
You have to realise these dimensions: 88% of all companies in the S&P500 of the largest US companies are worth less than OpenAI. This list from stocksguide.com includes well-known companies such as Starbucks, UPS, Nike, BlackRock and Goldman Sachs.
The valuation of over 150 billion USD is almost double the valuation in February 2024 (just 8 months ago) and ten times the valuation of a funding round 3 years ago.
That sounds crazy. And maybe it is. At least if you approach the matter with the usual mindset of an investor who is eagerly awaiting the IPO of his pre-IPO investment in order to seek a lucrative exit.
But perhaps the interests of the investors involved are quite different? Who are these investors willing to pay this truly strategic price to get a small piece of OpenAI?
It's worth mentioning that the official funding announcement on the OpenAI blog doesn't include the names of the investors.
OpenAI's investors
However, it is slowly becoming clear who the investors are and what concessions OpenAI had to make to get the money.
Let’s summarize the current state of information and speculation:
Investors in OpenAI's current funding round include a number of VC investors such as Thrive Capital, Khosla Ventures and Altimeter Capital.
Strategic partners Microsoft and Nvidia are also among the investors.
Softbank, ARK Invest and the MGX Fund are also participating.
It is worth noting that the company issued convertible bonds rather than traditional equity in its most recent financing. These must be repaid plus 9% interest if OpenAI does not succeed in abandoning its non-profit structure within 2 years. In the current organization, for example, there is a profit restriction for investors (including Microsoft), which is to be removed in the future.
Interestingly, OpenAI demands exclusivity from investors, i.e. they should not support VC-funded competitors such as Anthropic, Elon Musk's xAI, the search engine Perplexity and Safe Superintelligence by OpenAI co-founder Ilya Sutskever.
What will investors get for their money?
OpenAI is expected to generate 3.6 billion USD in revenue in 2024, with rising losses of over 5 billion USD. According to Reuters, investors were promised a further jump in sales to 11.6 billion USD in 2025. In 2026, this figure is expected to reach 25 billion USD, and in 2029, sales are expected to reach 100 billion USD.
This means that OpenAI is valued at 13-14 times next year's expected revenue in the latest funding round. This revenue multiple seems reasonable enough, it's in line with what you'd expect for a high growth tech company that's doing extremely well. Especially since ChatGPT alone now has 250 million weekly active users and the ChatGPT brand is synonymous with GenAI for many users.
So what is the problem? The creators of ChatGPT are currently burning through about 5 billion USD a year, and the ever-growing LLMs are causing exponentially rising costs in training and operations. In other words: In just one year, the next capital injection will be needed, and further dilution can be expected. The break-even point is still a long way off.
But Sam Altman did not get the money at this valuation because of ChatGPT's great success. His ambitions go far beyond Generative AI. He continues to entice his investors with the prospect of Artificial General Intelligence (AGI), which is supposed to develop cognitive abilities and be superior to human intelligence in the future.
I don't think this is realistic, because I don't think the current approaches to generative AI, with their exponentially increasing demands on computing power and energy, are capable of making this further quantum leap to AGI. Learn more in Will Locket's article on AI hitting a hard ceiling it can’t pass.
The Exodus at OpenAI
I am somewhat surprised that Sam Altman has successfully closed this round. Especially in light of the fact that he has seen a massive exodus from his technology management team in recent months.
Here is a summary of the departures just over the past few months:
September 2024
Mira Murati - Chief Technical Officer
Bob Mc Grew - Chief Research Officer
Barret Zoph - VP Post Training, started a new startup
August 2024
John Schulman, co-founder, now at direct competitor Anthropic, founded by former OpenAI employees and backed by Amazon and others
Greg Brockman, co-founder, currently on sabbatical (planned until end of 2024)
May 2024
Ilya Sutskever - Co-founder, Chief Scientist, launches new startup Safe Superintelligence
This means that 9 of the original 11 co-founders have left OpenAI. Apart from Sam Altman, only Wojciech Zaremba is still working for the company. This means that not only are many of the key figures in OpenAI research gone, but they are also hard at work building competing companies to OpenAI with billions in funding.
The AI Reality Check
I don't want to diminish the achievements and potential of OpenAI: Generative AI is on the verge of transforming the world of knowledge workers on this planet. This technological leap is probably only comparable to the advent of the Internet or the invention of the smartphone.
But I think all the predictions that AI will make the majority of people in the office obsolete in a few years are greatly exaggerated. Generative AI will help bring better and better tools and "co-pilots" to the market. But it will not take the thinking out of people's hands. Gen AI is not robust per se and will never be able to be used where humans need to reliably make the right small or big decisions.
Is OpenAI worth 150 billion USD today, given that the next dilution for shareholders is coming in about 12 months?
I don't think so, because I don't think OpenAI will make any more breakthroughs toward true AGI. I think that will require completely different approaches outside of or in addition to Generative AI. Whether OpenAI or another player will make these breakthroughs is completely open.
A company that "only" dominates the market for GenAI could well be worth a double-digit sales multiple despite a cash burn rate of over 100% if the business model is scalable, i.e. cost ratios fall and profits rise with more and more users.
For example, in any "normal" software company, at some point in the company's lifecycle, the cost of development and the cost of marketing and sales decrease. How this will work for OpenAI is beyond me, as training and operating increasingly complex LLMs consumes more and more resources. OpenAI - unlike other SaaS companies - has an incredibly capital-intensive and therefore unattractive business model.
Sam Altman himself has said that he needs 100 billion USD to achieve his goals. After this round, he has only raised 20 percent of that.
In addition, OpenAI faces stiff competition from equally deep-pocketed competitors in the field of Gen AI. First and foremost, of course, is Google. Then there is Elon Musk's xAI. And last but not least, Anthropic, founded by former OpenAI people and backed by Amazon. In short, it is unlikely that a monopoly-like market will emerge in the long term, where dream margins like those currently enjoyed by Nvidia can be earned.
The Interests of OpenAI Investors
So why are all these big-name investors putting billions more into OpenAI?
It's possible they just have more imagination than I do.
Or maybe they all have their own interests, according to which a failure of OpenAI would be much worse for them than risking an unprofitable investment.
My attempt at an explanation:
Microsoft
Microsoft, as the most important strategic partner of OpenAI, is to some extent dependent on the ChatGPT maker. The MS growth story for the next few years is essentially based on increasing revenues from the various copilots in Office and GitHub, as well as the AI workloads that enterprise customers host in the Azure cloud. In addition, an increasing valuation of OpenAI will have an impact on the value of the OpenAI stake, which was acquired at a bargain price from today's perspective in 2019 and 2023 (the valuation at the time was less than 30 billion USD). Microsoft will do everything it can to strengthen OpenAI.
Nvidia
Nvidia is a new OpenAI shareholder. This isn't a surprise, but it's right in line with Nvidia's playbook, which includes investing in 38 money-losing startups using Nvidia GPUs in 2023 alone. A few months ago, I wrote about what I consider to be Nvidia's (and Microsoft's) highly problematic role in the AI startup world: Big Tech as VC Investors: More red flags than laudable deals.
I figure that some of Microsoft's and Nvidia's investment in OpenAI will come back to them as revenue from OpenAI as their customer. There were probably at least revenue guarantees, as documented in previous cases. These relationships would only be revealed by an OpenAI IPO prospect, if it ever came to an IPO.
Venture Capital Investors
The VCs who have already invested are playing this game with ever-increasing valuations as long as they can raise more and more money for their new funds with the AI trend. They are primarily interested in increasing the valuation of their portfolio companies. And that's why they would give OpenAI even more money at a valuation of 300 billion USD, i.e. another doubling in 12 months. This will let them write up an existing holding and thus show a high book profit, which will help them raise new capital.
What is really interesting is which investors joines in the latest round: ARK Invest from Cathie Wood, Softbank and MGX, a sovereign wealth fund from the United Arab Emirates. Each of you can judge for yourself what you think of these investors.
I, personally, have not thought very positive of them in the past.
Perhaps even more interesting is which potential investors have said NO to a possible investment:
First, there's Apple. They were in talks for a long time and could have strengthened their existing partnership with OpenAI, but decided against investing. Sequoia Capital, which has become a critical voice in the VC world in the wake of the AI hype, is also not involved (see: When will the AI bubble burst?).
Is the OpenAI IPO coming?
The exciting question is what will happen next year when OpenAI needs to raise money again. If the valuation can be doubled to 300 billion USD, that won't be a problem, all the existing investors will be happy to come back. Microsoft and Nvidia for the usual reasons, the VCs are happy about the write-up of their portfolios and everyone is happy.
But what if the mood changes and no new investors can be found because, for example, OpenAI revenues aren't growing as fast as they once were, or people realize that we're not going to get real AGI anytime soon?
Then OpenAI could implode very quickly in the face of uncontrollable cash burn.
It would not be the first time that venture capitalists have had to explain to their clients why they have to write down their investments after a hype. For them, this is business as usual. The risk of not investing in a disruptive innovation is more serious. It's just venture capital. For other OpenAI investors, the crash is likely to be more painful.
It would be particularly bad if millions of retail investors ended up with OpenAI shares in their portfolios. But for that to happen, OpenAI would first have to pull off by far the biggest tech IPO the world has ever seen. I am sceptical that this will ever happen. But let's wait and see, with Sam Altman and new CFO Sarah Friar, the company has two of Silicon Valley's best storytellers in its ranks.
However, I suspect that Sam Altman will eventually realise that AGI is out of reach for the foreseeable future. In that case, there will probably never be an OpenAI IPO, but rather a takeover by, say, Microsoft. Once the AGI fantasy bursts, those technology stocks that have benefited from the AI hype, such as Nvidia, are likely to plummet.
In an interview with Bloomberg, renowned MIT professor Daron Acemoglu is even more pessimistic than I am about the consequences of the AI hype. He fears a crash and far-reaching negative economic consequences for the entire economy if the possibilities of AI continue to be seen too positively and too much investment is channelled into this area. A successful mega-IPO of OpenAI would probably mean the realisation of this negative scenario. I really hope it doesn't come to that.
Conclusion
The AI hype is entering the next round with OpenAI's huge funding round. At 157 billion USD, the company seems to me to be significantly overvalued. I see a big risk that OpenAI will go bankrupt under Sam Altman in the next few years, or be saved from bankruptcy by a takeover by Microsoft - at a much lower valuation.
If you are interested in my views on tech stocks, you can
*Disclaimer: The author and/or associated persons or companies do NOT own shares in OpenAI, Nvidia or Microsoft. This article is an expression of opinion and does not constitute investment advice.