LLAMA2 is out and it’s pretty fancy, and there’s a news report that they’re making a way bigger one. But if it’s just open source stuff, how do they justify the massive costs to make it? It seems like everyone can just use it for free.
In: 156
I’m not certain if this is their strategy, but this is an obvious one:
First one to Artificial General Intelligence (AGI) wins the world.
If we can come up with a problem-solving reason/act/observe AI that can solve any problem presented at a human level, that is called AGI. If you were talking to it via chat, it would be indistinguishable from a human with an internet connection, except faster. If you gave it a complex problem and gave it 2 days to run, you’d come back to find it has accomplished as much as a competent human would have in 6-8 days (because it never sleeps or takes a break). This is a world-altering technology. Imagine every task currently assigned to a human at a keyboard can now be automated.
If you’re the first one there and you can patent or close-source the secret sauce that makes it work, that’s a multi-trillion dollar asset you’ve built.
Then, you could instruct it to review its own code and make itself smarter, and in a few months it could possibly produce an ASI: Artificial Super Intelligence, an intelligence beyond human levels. This is an event known (controversially!) in tech circles as “the singularity”. If an AGI would put a huge number of humans out of work, an ASI could transform society and redefine what it means to be human. A couple of decades with an ASI and the projections start to sound crazy.
Spending a few billion now to get to AGI later, and maybe wind up with an ASI under your control is possibly the best investment in the history of the world.
Latest Answers