Chronicles of 2025

Lucky I don’t have that problem… they all are too expensive for me. Or, in other words, I did not catch them on time with one exception: Supermicro. Bought at split-adjusted $4.32 back in 2022 when they were valued as “just another hardware screwdriver driver”. Then I bought some more the same year at $6.42, still cheap after a 50% price appreciation.

Big volatility due to some (not confirmed) bookkeeping falsifications. Of course my method made me cash out already many multiples of my investments, but at the moment I probably would buy some more, they are not very expensive fundamentally. But I do what my capt’n tells me to… and this is to just hold on to my investment.

Where do you think Oracle is going to get their chips from?

1 Like

Intel?

1 Like

If only they had an AI product before 2027. AMD seems to be an actual second source supplier of AI accelerators, but I’d guess the most is Nvidia again.

Intel bought Habana labs in 2019 and went on to destroy their promising line-up of AI accelerators. They actually bought and destroyed several AI hardware companies before ChatGPT was released.

1 Like

The Wall Street Journal has a nice article on this.

Excerpt (highlighting mine):

The OpenAI and Oracle contract, which starts in 2027, is a risky gamble for both companies. OpenAI is a money-losing startup that disclosed in June it was generating roughly $10 billion in annual revenue—less than one-fifth of the $60 billion it will have to pay on average every year. Oracle is concentrating a large chunk of its future revenue on one customer—and will likely have to take on debt to buy the AI chips needed to power the data centers.

Guys, I was being sarcastic. :wink:

4 Likes

Great, this gives us a nice run of hype to 2028 then! /s? I am not sure myself!

1 Like

Does anyone else wonder whether it is time to dig out this one, anno 1995:

4 Likes

Did I get this right? :innocent:

/s
6 Likes

I think we’re pretty close to the peak now. Next stop: trough of dillusionment.

3 Likes

Mad PowerPoint Skillz at play (ok I don’t have a lot urgent to do at work and on holiday next week so brain is switched off)

6 Likes

like on a techno party… waiting for the drop! :mirror_ball:

3 Likes

This is where my strategy comes into play.
I always sell high and buy low.

6 Likes

or a mosh pit wall of death, waiting to close in…

Pessimist.

(Isn't that my role?)

Things aren’t dire yet.


Edit: Chewbacca there in the background, is he $MSFT …? Or $META? Maybe $GOOG?

The plateau is going to be 3-4 orders of magnitude higher in a decade than either the peak or the trough, so it doesn’t matter where we are today.

I’m kidding

Or am I?!

1 Like

Probably not, a plateau is not necessarily a bad thing as long as the holy dividends are rolling in.

My company, who doesn’t pursue anything to results, is gonna invest in AI because some marketer talked to them. Guys/gals, run for the exits.

5 Likes

The lost decade 2000-2009 would like a word.

expecting 3-4 magnitudes in a decade is extremely overly optimistic.

A rational expectation is 5%/year real after inflation return. That is the long running average over hundreds of years by now.

Also current valuations would actually give higher chance to quite depressed returns over the next decade. Not that it would predict it for sure. Just that the past stats would give that expectation.

But it‘s really not a stretch to say there is a solid chance to have exactly zero (real) return over a decade. That has happened a lot of times already.

3 Likes

It’s actually extremely pessimistic.

The S&P500 is currently just under 6’600.

So if we take 3-4 OOM that goes to: 6’600’000 to 66’000’000.

You don’t get that through growth. You get that through hyperinflation and currency debasement.

5 Likes