OpenAI & Microsoft – ChatCopilot

OpenAI bites the hand that feeds it. 

  • Despite the inevitability of ending up as part of Microsoft, Open AI is determined to compete with Copilot and is making acquisitions to bolster its service to prevent early subscribers from defecting to Microsoft’s Copilot.
  • OpenAI was first out of the gate with an enterprise offering by adapting ChatGPT to have a large context window and faster response time and in the hype, it has signed up 150,000 users at $30 per user per month.
  • PwC adding another 100,000 brings that total to 250,000 creating a business with $90m of revenue.
  • ChatGPT Enterprise is not much more than a souped version of the consumer offering which leaves it vulnerable to competition from Microsoft.
  • Microsoft’s history as a supplier of enterprise software means that it has a lot more bells and whistles that it can pack around its generative AI offering that will allow its functionality to outstrip that of OpenAI,
  • This combined with the slowing of the rate of improvement of new models (see here) and a similar price leaves ChatGPT Enterprise looking pretty uncompetitive against Microsoft’s Copilot.
  • I think that this is why OpenAI is quietly acquiring the bits that it is missing and hopes to cobble them together into something more competitive.
  • For example, OpenAI has quietly purchased a collaboration technology system for Mac called Multi and last week acquired Rockset which is a small start-up focused on retrieval augmented generation (RAG).
  • Multi has been sunsetted with immediate effect and I assume that the collaboration technology will be woven into ChatGPT Enterprise.
  • This would enable the AI-team member that Microsoft has launched as well as allow collaboration within a team all within the OpenAI enterprise experience.
  • This will initially be focused on the Mac where there is currently no Copilot available outside of a web browser but if it becomes a very popular feature of Microsoft Office, I would not be surprised to see Microsoft port it properly to the Mac.
  • There is a Copilot app in the Mac Store but it looks to me to be almost the same as running it from the web browser meaning that there is no OS or app integration which is what is now beginning to be offered on CoPilot+ PCs.
  • This gives OpenAI a fairly clear run at the Mac and if it proves to be very popular, I would not be surprised to see it compete head-to-head with Microsoft on its turf.
  • However, whether OpenAI is still an independent company by that time remains to be seen.
  • OpenAI has not fixed the fundamental conflict which caused it to almost implode last year and, as such, this is still a bomb waiting to explode.
  • Satya Nadella has done a stellar job since he became CEO at Microsoft and has completely turned the corporate culture on its head and done so with fantastic effect.
  • However, the one blunder he has made is to create a single source dependency of most of his revenue-generating products where that source could self-immolate at any moment.
  • This is why I think that Microsoft has no choice but to buy OpenAI at some point.
  • This point is likely to be either the next time OpenAI sets itself on fire or when the AI bubble pops and everyone loses interest and cuts spending.
  • This is why when I think about the competition in AI, I think of Microsoft and OpenAI as one company despite OpenAI’s attempts to maintain its independence.
  • This acquisition looks inevitable to me.

RICHARD WINDSOR

Richard is founder, owner of research company, Radio Free Mobile. He has 16 years of experience working in sell side equity research. During his 11 year tenure at Nomura Securities, he focused on the equity coverage of the Global Technology sector.