AGI is an engineering problem, not a model training problem

Re:AGI is an engineering problem, not a model training problem by fuckaj on Saturday August 23, 2025 @11:07PM (#45000978)

The physical universe has much higher throughput and lower latency than our computer emulating a digital world.

[ Parent | on HN ]

Re:AGI is an engineering problem, not a model training problem by tomrod on Sunday August 24, 2025 @12:14AM (#45001341)

Wouldn't it be nice if LLMs emulated the real world!

They predict next likely text token. That we can do so much with that is an absolute testament to the brilliance of researchers, engineers, and product builders.

We are not yet creating a god in any sense.

[ Parent | on HN ]

Re:AGI is an engineering problem, not a model training problem by fuckaj on Sunday August 24, 2025 @01:02AM (#45001536)

I mean that the computing power available to evolution and biological processes for training is magnitudes higher than for an LLM.

[ Parent | on HN ]

Re:AGI is an engineering problem, not a model training problem by tomrod on Sunday August 24, 2025 @04:12PM (#45007359)

Is it? Seems like C. elegans does just fine with limited compute. Despite our inability to model it in OpenWorm.

[ Parent | on HN ]