Epoch AI measures how fast the cost of a given level of AI performance is falling across five benchmarks covering math, science and games of skill: about 47% per quarter, or 13x per year, since 2023, faster than electricity, compute, batteries or DNA sequencing ever fell.
It will have some uses. But it will not be the foundation of everything in future tech like they are claiming. It will be some limited mainstay markets, which doesn’t in any way justify the size of the hyperscaling they are trying to achieve. All that is pure posturing for the stock value.
Especially since the open source models may likely win out over the big expensive corpo models, based purely on costs.
So what about 5 years? Forget the LLMs, seemingly alot of positive things are happening in various industries thanks to advancements in AI (machine learning, robotics, etc). What about 10 years? What about 25? This is one of those things that will keep advancing, just like computers themselves.
It will have some uses. But it will not be the foundation of everything in future tech like they are claiming. It will be some limited mainstay markets, which doesn’t in any way justify the size of the hyperscaling they are trying to achieve. All that is pure posturing for the stock value.
Especially since the open source models may likely win out over the big expensive corpo models, based purely on costs.
So what about 5 years? Forget the LLMs, seemingly alot of positive things are happening in various industries thanks to advancements in AI (machine learning, robotics, etc). What about 10 years? What about 25? This is one of those things that will keep advancing, just like computers themselves.