7 Comments
User's avatar
Ngozi's avatar

Hi Azeem, interesting breakdown but I was expecting more around the subheader “How i think about complex problems”. I think you demonstrated that its first principles for the most part but would be interested in more on that topic

James's avatar

Don’t forget the Jony Ivey angle.

Azeem Azhar's avatar

maybe. but there is zero evidence we will get a product that anyone likes. so i would discount that to zero.

Will S Johnston's avatar

What's the moat that OpenAI has? I don't see it. It seems that LLMs are becoming commoditized and I suspect that Google will take the lead because they have so many others applications which are in our daily lives.

Azeem Azhar's avatar

Moats aren’t as simple as that, of course. Perhaps owning the the term is one. Perhaps have power relationships in multiple markets could be an advantage for a few years.

Will S Johnston's avatar

I worked at Inktomi ($1b market cap) when there were a range of popular search engines (1999) and no one was dominant or had a moat. Suddenly Googles search engine was far outpacing the rest and they sewed it up. They had a moat which was providing by far the best algorithm. All of the other well capitalized search engines are now in the graveyard. I see a lot of parallels.

Azeem Azhar's avatar

I did a deal with Inktomi back in 1999 for a UK TV company. I remember those days well.

There is somethng more a moat for OpenAI -- user memories; scheduled tasks and background agent work; plus the enterprise business is a different kettle of fish