One thing about these numbers that's absolutely shocking to me is how low the energy use is:
> That model’s usage was well within our budget ($68, about 4kWh of energy use / 365 grams of carbon emissions).
The energy cost is literally 1% of the total cost. For context, 4kWh of energy would drive you about 15 miles in an EV, about half of the average person's daily driving miles. It's boiling 10 gallons of water.
With the talk of AI data centers' impact on the world, you'd think this would be 10x to 100x the amount of energy in order to get the effects they're using here.
My takeaway: the AI data center buildout is an overbuild probably at least as large as the fiber buildout that left us with so much dark fiber. If not even bigger. The only thing that will save the economy is the inability of NVIDIA and chip fabs to produce enough chips to match the buildout planned.
> With the talk of AI data centers' impact on the world, you'd think this would be 10x to 100x the amount of energy in order to get the effects they're using here.
This was going so well until this. Everything at scale has environmental impact because you centralize the downside and distribute the upside. This is an important alienation, but it hides the amount of heat, noise and impact on distribution that datacenters have on local infrastructure and environment.
My point is the impact is real, measurable, but nobody will do it because nobody can stop the hype train lest they be labeled opponents to the progress
These days even the AI companies are urging a slowdown if pace.
If somebody is afraid of speaking their mind, because the are afraid of being labelled an opponent to progress, that's some personal issue to work through.
It is popular and encouraged to be skeptical of AI, there no social opprobrium about it, unless you're in an extremist political cult, in which case you probably call it SI.
You can't really say this categorically, because that's not a universal experience.
Maybe you are one of the lucky ones not having immense pressure at work to adopt AI at all costs, but the reality of it is that if I open my mike in a "debate" where I work with all devs and managers, to speak in favor of taking things easy and to give time for people to learn these tools properly, I'd be frowned upon.
So while it's true that's encouraged to be skeptical about the tech, this really depends on the context and saying it doesn't clashes hard with my experience of the world.
epistasis · · focus · HN ↗
> That model’s usage was well within our budget ($68, about 4kWh of energy use / 365 grams of carbon emissions).
The energy cost is literally 1% of the total cost. For context, 4kWh of energy would drive you about 15 miles in an EV, about half of the average person's daily driving miles. It's boiling 10 gallons of water.
With the talk of AI data centers' impact on the world, you'd think this would be 10x to 100x the amount of energy in order to get the effects they're using here.
My takeaway: the AI data center buildout is an overbuild probably at least as large as the fiber buildout that left us with so much dark fiber. If not even bigger. The only thing that will save the economy is the inability of NVIDIA and chip fabs to produce enough chips to match the buildout planned.
gchamonlive · · focus · HN ↗
This was going so well until this. Everything at scale has environmental impact because you centralize the downside and distribute the upside. This is an important alienation, but it hides the amount of heat, noise and impact on distribution that datacenters have on local infrastructure and environment.
epistasis · · focus · HN ↗
Calculate the scale. Add it up. Do it.
Look at the numbers and come back to me.
gchamonlive · · focus · HN ↗
epistasis · · focus · HN ↗
If somebody is afraid of speaking their mind, because the are afraid of being labelled an opponent to progress, that's some personal issue to work through.
It is popular and encouraged to be skeptical of AI, there no social opprobrium about it, unless you're in an extremist political cult, in which case you probably call it SI.
gchamonlive · · focus · HN ↗
You can't really say this categorically, because that's not a universal experience.
Maybe you are one of the lucky ones not having immense pressure at work to adopt AI at all costs, but the reality of it is that if I open my mike in a "debate" where I work with all devs and managers, to speak in favor of taking things easy and to give time for people to learn these tools properly, I'd be frowned upon.
So while it's true that's encouraged to be skeptical about the tech, this really depends on the context and saying it doesn't clashes hard with my experience of the world.