Nvidia CEO Jensen Huang — whose work helped propel artificial intelligence — is stressing in an Associated Press interview that society has no choice but to change in the advent of AI.
I remember seeing videos in the early 80s when US federal seatbelt requirements went into effect and everyone in an uproar about it. Screaming government overreach, the nanny state, etc. Now anyone not wearing a seatbelt is an idiot.
AI will end up being a useful utility and those opposing it the loudest now will no longer be loud about it. Right now it’s expensive, it’s inefficient, it uses too many resources much like cars from the 80s compared to cars now. As Moore’s Law kicks in it will use fewer resources, become more efficient, and more accurate
If you compare similar models, yeah. A modern Corolla is like 500 pounds heavier. Let’s not get started on trucks, or the fact that most cars are now SUVs.
Tell me you don’t know how the technology works without telling me.
The “good” models are huge. They make servers larger than your local town centre run on full load for every query. There’s no throttle to hold them back, as that doesn’t help. It’ll always be full speed ahead. And the results are based on probability, with a factor of bullshit baked in as part of the design. Yes, computers become more efficient over time, but that doesn’t help when you keep pushing them to go full speed all the time.
The thirstiest CPU in the original Pentium family used to consume 17.3 Watts at full speed. Today, the Intel Xeon w9-3595X slurps 462 Watts given the chance. The Nvidia B200 datacenter GPU has a TDP of 1000 Watts. And data centers have thousands of these running, full speed, for LLMs to serve us hallucinated “facts”.
So AI of the future will be more useful in more cases and use less energy and other resources. Based on… nothing within the AI technology, but on a very loose analogy.
There are very probably architectural limitations which will prevent large language models from ever getting much better than they are right now. They are most likely a dead end.
Seatbelts in cars actually do what they say on the tin. LLMs have some uses but for most things they don’t do what they say on the tin. AI boosters and bad comparisons. Can’t name a better duo
I remember seeing videos in the early 80s when US federal seatbelt requirements went into effect and everyone in an uproar about it. Screaming government overreach, the nanny state, etc. Now anyone not wearing a seatbelt is an idiot.
AI will end up being a useful utility and those opposing it the loudest now will no longer be loud about it. Right now it’s expensive, it’s inefficient, it uses too many resources much like cars from the 80s compared to cars now. As Moore’s Law kicks in it will use fewer resources, become more efficient, and more accurate
Cars now use too many resources.
Compared to 30 years ago?
If you compare similar models, yeah. A modern Corolla is like 500 pounds heavier. Let’s not get started on trucks, or the fact that most cars are now SUVs.
Plus there’s a hell of a lot more of them.
And more roads. And more parking lots.
I’d much prefer reliable, clean public transit, and it’s entirely doable. But better cars too because people.
Tell me you don’t know how the technology works without telling me. The “good” models are huge. They make servers larger than your local town centre run on full load for every query. There’s no throttle to hold them back, as that doesn’t help. It’ll always be full speed ahead. And the results are based on probability, with a factor of bullshit baked in as part of the design. Yes, computers become more efficient over time, but that doesn’t help when you keep pushing them to go full speed all the time. The thirstiest CPU in the original Pentium family used to consume 17.3 Watts at full speed. Today, the Intel Xeon w9-3595X slurps 462 Watts given the chance. The Nvidia B200 datacenter GPU has a TDP of 1000 Watts. And data centers have thousands of these running, full speed, for LLMs to serve us hallucinated “facts”.
So AI of the future will be more useful in more cases and use less energy and other resources. Based on… nothing within the AI technology, but on a very loose analogy.
There are very probably architectural limitations which will prevent large language models from ever getting much better than they are right now. They are most likely a dead end.
Seatbelts in cars actually do what they say on the tin. LLMs have some uses but for most things they don’t do what they say on the tin. AI boosters and bad comparisons. Can’t name a better duo