Hard questions for AI investors after OpenAI drama
For technology investors who were looking to hitch their wagon to OpenAI and its world-beating generative AI technology, the boardroom events of the past weeks have given some reason to pause.
It’s only been a few weeks since PitchBook reported $US21.5 billion had been poured into AI start-ups this year — four times the value seen in 2022 — but OpenAI’s decision has the potential to slow the flow.
The story continues to shift as new events unfold, but OpenAI’s non-profit board abruptly sacked Sam Altman, who has become the face of AI through the company’s dominance in ChatGPT and image generator Dall-e 3.
Days after a victory lap at the OpenAI Development Day, which saw so many new people rush to the paid version of the software that the company had to pause sign-ups, Altman was ousted Friday in the US.
Chaos immediately ensued, with key OpenAI investor Microsoft initially encouraging Altman to be rehired, then hiring him themselves — and in doing so, opening the door for what could be hundreds of disaffected OpenAI employees to jump ship – which then ultimately forced out the OpenAI board and prompted Altman’s rehiring.
The impact on Microsoft shares, owner of 49% of OpenAI’s for-profit arm, was interesting to watch. Shares fell late Friday, then recovered to hit an all-time high, before moderating mid-week while the market frantically tried to parse the hour-by-hour changes.
Even more interesting, though, was the impact on OpenAI’s (unlisted) equity.
What does the drama mean for Aussie tech investors?
Secondary buyer interest in $100 million of OpenAI’s stock was said to have “vanished” as investors waited to see what would eventuate, while venture capitalists — already wary of investing in software that did little more than wrap ChatGPT functionality — saw their fears about learning on OpenAI technology realised.
So what does that mean for Australian tech investors who have a strong appetite for AI exposure?
Let’s go back to basics on what AI investment means, which we can view through three lenses.
- The first investment path gives you exposure directly to the giants, Microsoft, Nvidia and others, or the infrastructure (like data centres and chips).
- The second path is economic — tapping into the general productivity benefits that can be reaped through investment in AI processes, technologies or partnerships. What profit margin improvements are there to be had from mature businesses with large numbers of employees?
- The third is at the business level, looking at those companies that are able to add value in the AI ecosystem, by using AI or by monetising their proprietary data in some way.
All three levels have a value for tech investment, but the risk profile varies.
If you consider the winners and losers of the past few days of activity, Microsoft appears to have come out ahead. In the final shake up, while it appears the only change is to OpenAI’s board, it has highlighted three important things.
Firstly, there is the incredible loyalty shown by the staff to Sam Altman, next, the strength of the relationship between Sam Altman and Microsoft, which ultimately improves their credentials, and finally the management of an unforeseen risk to Microsoft’s investment, which they have now been able to address through influence on the new board appointees.
Microsoft competitors such as Amazon Web Services (partnering with Anthropic) and Google (still trying to produce a suitably impressive GenAI service) could have benefited from an OpenAI break-up. This no longer looks like a prospect.
In the losing camp are those AI wrapper companies that have leaned heavily on OpenAI’s general approach to sharing their large language models. In an alternative world, OpenAI shuts down which exposes significant business risk.
In a world of ‘move fast and break things’, many companies, from the second largest in the world to the smallest start-up, just had a nasty peek into what “break” looks like. Fortunately, in this case, it was just a scare.
Much will also now ride on whether Copilot is as impressive as forecast as stumbling now would be a critical threat not only to Microsoft but also to the adoption of generative AI overall.
There’s much unknown but one thing remains certain.
If finding the right AI investment was challenging before, things have become significantly harder.