Re:AGI is an engineering problem, not a model training problem
by energy123 on Saturday August 23, 2025 @11:22PM
(#45001061)
What will it scale up to if not AGI? OpenAI has a synthetic data flywheel. What are the asymptotics of this flywheel assuming no qualitative additional breakthrough?
[ Parent
| on HN ]
Re:AGI is an engineering problem, not a model training problem
by supermatt on Saturday August 23, 2025 @11:29PM
(#45001090)
What will shouting louder achieve if not wisdom?
Re:AGI is an engineering problem, not a model training problem
by richyg840 on Saturday August 23, 2025 @11:43PM
(#45001177)
It's established science (see Chomsky) that a probablistic model will never achieve something approaching AGI
Re:AGI is an engineering problem, not a model training problem
by HarHarVeryFunny on Sunday August 24, 2025 @01:49PM
(#45006188)
Did GPT-1 scale up to be a database ? No - it scaled up to be GPT-2
Did GPT-2 scale up to be an expert system ? No - it scaled up to be GPT-3
..
Did GPT-4 scale up to become AGI ? No - it scaled up to be GPT-5
Moreover, the differences between each new version are becoming increasingly less. We're reaching an asymptote because the more data you've trained on, natural or synthetic, the less is the impact of any incremental additions.
If you scale up an LLM big enough, then essentially what you'll get is GPT-5.