Comment by josu
11 days ago
> invested in (...) anything else practical.
I don't understand how this is the top comment. LLMs have unlocked a lot of value for me personally, and arguably for the society as a whole. They are also one of the coolest technologies I've tried in years. As a technologist, I'm really glad that money is pouring in and allowing us to find its limits.
Not sure if this is the same sentiment, but I feel that even though the technology is handy, the investment is speculation. Billions and billions are being spent on developing this technology hoping to create a unicorn that turns billions into trillions.
I know it's unrealistic, but imagine you took $30b and just started ploughing into a random city. Real estate, development grants, EV charging infra, renewable projects, transit. Just boring existing technology, but concentrated investment to make the other investments more valuable. You could turn that $30b into $60b in a few years while changing the lives of an entire city.
But American capital says no, and gambles it, because it's just a game to the people in charge.
Money is responsibility. While I pinch my pennies and show up to my day job, the people with the keys throw billions on an oil fire hoping it magically turns into gold.
> LLMs have unlocked a lot of value for me personally, and arguably for the society as a whole.
The second part doesn't follow from the first which, in turn, is irrelevant here because we aren't talking about any value for society, we're trying to figure out if that value is greater than its opportunity cost.
The "AI" has been with us for several years now but it hasn't improved overall productivity in any meaningful way and while it's use is already plateauing it keeps sucking in capital like a giant black hole.
> As a technologist, I'm really glad that money is pouring in and allowing us to find its limits.
That's not a good thing when it happens at the expense of the rest of society.
It's only natural that workers at NVIDIA and other suppliers of AI-related products would like the gravy train to travel their way in perpetuity, but they are literally wasting capital, which isn't theirs, for their own enrichment, while creating inflation, shortages and misallocations on a grand scale.
The issue isn't whether LLMs are useful: it's whether OpenAI is positioned to deliver substantial future profitability.
Given their competition consists of the largest tech companies in the world, several of which are vertically-integrated, have advanced AI research programs, and own the last hardware mile to the customer...
OpenAI better hope it can maintain a performance lead, while being able to run inference cheaper than its competitors, while also launching a must-have device it controls (AR/audio glasses?).