“This doesn’t make sense. We’ve only added 200ppm of CO2 to our atmosphere. Shouldn’t affect anything.”
Systems at a roughly stable equilibrium can be surprisingly easy to shift out of that equilibrium by pushing them ever so slightly outside a local minima. Nobody right now is saying this particular situation is going to result in catastrophe, but we should exercise some caution when we are causing observable effects to the one planet we inhabit.
The person who briefed me on the change in prediction of global climatic modeling, pushing the recovery back 75 years, also said this:
"If you are in a flying 747 flying directly at another one, then it appears as a very very tiny spec in the sky, until you get close enough to see it, and then it gets really really big, really really fast, and then you are dead."
He also said this about stable equilibrium: You can roll a glass in a circle on a table, and it can roll in a circle for a long time, but if it falls off the table, it is going to take some effort to get back to equilibrium.
We have little idea what equilibrium is, and we also have less idea about what it will take to get back to it, after leaving it.
If you take a basketball, and breathe on it, the frost collected on it, is less than the thickness of the atmosphere. Many phenomenon are even less detectable, but see the prior comment about 747s.
Sorry for the late reply. I was being somewhat sarcastic, or maybe not. During the "dot com" era, investors were throwing money at anybody willing to take it, and a start-up was questioned about whether it was maintaining a sufficient "burn rate." In hindsight, it was a speculative bubble.
The company that I work for survived that era by not doing those things.
I guess, but if you invested $1 million into 100 companies in 2000 and one of them was Amazon (and other 99 went to zero) your investments combined would have had an annual growth of 24% over the past 25 years.
Someone in 2000 would not have done your proposed strategy unless they could name 100 .com like tech companies from 1975 that, if you had invested $1 million dollars in, would have returned 24% over the previous 25 years.
If the idea is just look at what worked 25 years ago, the 2000 investor would probably have followed a different approach.
Why should they be in that market at all? I don't understand why we've got to this place where every big tech company needs to have it's fingers in every pie or "they're falling behind". AI / LLMs look to be on their way to being infrastructure more than product in and of themselves. A patient huge company could build on top of what others are doing and wait for the market to settle, then buy for vertical integration.
They're not even in the same ballpark as the SOTA models, and they're not even in the same ballpark as the SOTA open source models, which was supposed to be their niche.
But even if they were, it's not immediately clear how they plan to make any money with having an open source model. So far, their applications of their AI, i.e. fake AI characters within their social media, are some of the dumbest ideas I've ever heard of.
Deepseek is a solid competitor in their niche, the open-source LLM niche. Llama 4 didn't impress, and so they are very vulnerable even in the free niche they carved out for themselves
How would you even begin to calculate that? They use AI to make their products more valuable to advertisers, but there's no way to say how much their revenue would decline if they suddenly eliminated AI.
reply