Sam Altman vs Ilya Sutskever: What is happening at OpenAI? And what is Microsoft doing?

I’m writing this at 21:12 UK time. 13:12 Pacific Standard time over in San Francisco. As of right now Sam Altman is still out of OpenAI as CEO, but may be coming back (again) after co-founder Ilya Sutskever seemingly changed his mind on the ousting he might have started. This post isn’t a live commentary, for that check out the excellent coverage at The Verge. Instead I want to make some sense of what is going on, for us in marketing looking at AI as an accelerator to our goals.

Who is Sam Altman? What happened?

I’ll keep this bit short. It is based on facts with a touch of opinion. The fact bit first. Sam Altman was one of the founders of OpenAI and it’s CEO. He is a darling of tech, very successful and the leading light in AI. Ilya Sutskever co-founded Open AI and is Chief Scientist at the company. He seems to have asked the OpenAI board to remove CEO Sam Altman. His very recent tweet of regret backs this up. The board did so with a very strong statement. This was the best bit:

The rest gets a bit muddled. It seems Ilya wanted Sam out because he felt Sam was too focused on commercialising the tech at the heart of the company, which would comprise the true vision of OpenAI. We will come back to that vision in a bit. Sam was maybe looking at setting up other (for profit) companies to complement OpenAI.

Sam Altman, CEO of OpenAI (Image credit: James Tamim)

It is also thought the Nov 6th OpenAI Dev day was taken as a sign that Sam was looking at how OpenAI could start making money (something very much for profit Microsoft, who own 49% of OpenAI, might be keen on) faster than others in the business were comfy with.

OpenAI started as a non-profit. With a vision to do something very big (see next bit). Then Sam changed that so companies like Microsoft would be more likely to invest. And invest they did, a reported $10 billion. Most of which is spent on Azure services to power the models at the heart of OpenAI. You see AI is an eye wateringly expensive business to run.

Let’s talk about Artificial General Intelligence

This is were things get really interesting. You see OpenAI wasn’t created to give us the ChatGPT, or a means to make pictures with Dall-E. Nor write code, or speak to us (try the mobile app if you haven’t). No, it was created with this single purpose:

And we all know what Artificial General Intelligence is? AGI? No? Ok buckle in. AGI is the holy grail of AI research, something that could revolutionize the world as we know it. Imagine an AI that’s not just good at one thing (generating pictures of cats, writing social copy), but has the intellectual prowess of a human, capable of understanding, learning, and innovating across any field – from composing symphonies to solving climate change. And then going beyond that, at exponential rates. AGI won’t just be a tool; it could be the dawn of a new era in technology, where AI could think, adapt, and create far in excess of humans.

Many smart people think it will never happen, or will take 100’s of years to develop. But it doesn’t matter really. Sam and IIya have set up a company whose sole aim is to create it. Not to power tools like Jasper or Microsoft Copilot. Not to offer a chatbot to act as a personal friend, or improve healthcare. They are having a crack at building something that is smarter than all humans!

And one of them (Ilya) just fired the other one (Sam) became he thought he (Sam) was doing it too fast and risking the safety of all of us (there is a little conjecture here on my part, but only a little).

This is pretty sensational stuff. And it is playing out before our very eyes. OpenAI scientists are trying to create this tech right now. Even Sam himself thinks it might be really dangerous. Ilya said last year he thinks ChatGPT 4 (yep the one you play with right now) might be “slightly conscious”. I may never look at my ChatGPT Plus subscription in the same way again.

Back to planet Earth. Or maybe planet Microsoft

Ok let’s bring this back to normality a little. Us marketing folk are using AI every day. Fifty Five and Five are building tools on this tech that our clients are quickly relying on. So what does all this mean for us? Should we care? Should we think about AGI? Can we rely on AI to run our businesses and hit our marketing objectives?

The simple answer is yes. Because Microsoft.

As of the time of writing (now 22:46 UK) Sam Altman now works for Microsoft, heading up some hither unknown AI operation (basically they created something just for him this weekend). 500 of the 700 employees at OpenAI who are threatening to quit if Sam doesn’t return, have been offered jobs in said Microsoft division. And Sam has publicly said ‘he and Satya‘ won’t let OpenAI fail. And that is it in a nutshell.

Microsoft have put $10billion into OpenAI. They have bet their Microsoft 365 (and Windows) offering on the OpenAI powered Copilot. Everything runs on Azure. OpenAI is too big and too important for Microsoft to let it descend into full on chaos, let alone fail.

We might see OpenAI back to normal tomorrow, with Sam again CEO. Or a new version of the company inside Microsoft, with Sam and 500 former OpenAI engineers working like nothing happened. Its the same end result for us really.

(Oh and I get this might feel like a big capitalism fueled ‘huge tech giant from Redmond wins again’ situation. And it might be that. And that might have down sides (Microsoft has history here) that need to be looked at. It’s a topic for another post.)

OpenAI is going nowhere. It is too important to Microsoft for them to let that happen. It is a stable platform that will be with us in one form or another for a long time. Fifty Five and Five will be building on it, and we are confident it is the right choice for our clients.

Whether it becomes AGI or not, well that really is another blog post.