Big Tech muss einen Jahresumsatz von 600 Milliarden Dollar erwirtschaften, um Ausgaben für KI-Hardware zu rechtfertigen

How long until you see a major corporation announce they’re pulling back on their AI spend?

Is that a lot?

That sounds like a lot.

For context, google search is a ~$150 b/yr product. So, we just need to build four more Googles and we’re good.

Rip everyones job but the handy man I should’ve listened to south park.

In reality it needs probably 300 companies to find a 2B/year product for it. Like medical imaging results, or quickly analyzing people who come into the emergency room, call service centers become much much better, analyzing oil and gas fields, etc.

It really doesn’t require just 4 companies to come up with some massive solution.

It is also very dangerous with this type of tech to not be in a position to run with whatever someone else brings to market first. First to market in this big tech could mean several big tech companies lose all their competitiveness. The speed at which it’s advancing each year means someone can’t play 3 year catch up. Trying to hire people, trying to purchase equipment, then trying to play catch up, even a year might make it hard to catch up, especially if the other parties are sabotaging your efforts. If they hire up available talent to just block a company, or make such large hardware purchases that another company can’t acquire anything for several years, it means they could now be behind 7-10 years.

There are times to gamble, and this probably isn’t one of them.

That’s presuming there’s any goal to break even, which is entirely unnecessary in today’s never-ending “fundraising” business environment. In fact, Lyft, Pintrest, Zillow – all large, ongoing corporations without profits. Uber only posted its first profits last year, after 5 years as a *public* company.