This is unpopular in the current climate of hysteria, but it's worth reminding ourselves: data centers are probably one of the least polluting industrial sites imaginable. That's not to say it's nothing, but contrast it with, say, a refinery. Or mineral processing. Or basically any other type of industrial activity on which modern civilization relies.
The problem with your reasoning is that nothing ever matters, because you can always divide things in smaller and smaller chunks and in the end "it's really not that bad it's only 0.x percent of emissions/pollution/etc"
I agree; but that's also not a very useful observation. Modern society depends on industry of all types, and national independence depends on at least some of it being local. We can try to make it cleaner, and to some degree we can choose which types to implement. But we do not have the option to just wish it all away.
It would be very nice if we could have our cake and eat it too.
Would you mind a data center sprouting up in your backyard? Just ask residents near Elon's new gas turbine powered GPU datacenters how lovely they find the hum.
Framing the choice as a dichotomy between data centers and other industrial plants misses the obvious third option: nothing at all, preserve the status quo.
"playdates" is in itself a tell. That didn't used to be a thing. Kids would just... go out and find their friends and play. No parental scheduling or coordination required.
We use this term when little kids aren't old enough to just go out and find their friends and play. E.g., 2-5 year olds definitely aren't going to be let to go wander around the street.
Effectively nobody considering buying into the IPO cares about Anthropic's source control system. They could be passing around floppy disks and so long as they were still shipping nobody would care.
Not about the details but the marketing story is all about how "agents" are building so fast they need to design new infrastructure for the new world, etc. Extra points if you spin it as though the "agents" have built some new never-before-seen architecture that is clearly better than the "human scale" stuff.
The intended audience is obvious from first paragraph alone:
> At Anthropic, coding is no longer the bottleneck to building systems or doing research. Claude is able to autonomously iterate on research loops, micro-optimize complex programs, and build new services from the ground up at a speed and scale we haven’t seen before. The bottleneck, now, is all of the developer infrastructure built for human scale: source control, builds, and CI.
Translation:
> We are so far at the cutting edge we need to build shiny new infrastructure to replace the smelly "human scale" cruft. How could you /not/ value us at $2T?
Scaling what exactly? Anthropic doesn't release any new products except models. I'm sure you don't need tens of thousands of pushes every day to train models. Scaling fast is so far just their words, nothing else indicates it.
NVidia has enormous operating margins, so a competitive solution doesn't have to match or beat NVidia's scale efficiencies; it just has to beat delivered cost.
One objective of the project might be simply to provide credible negotiating leverage when dealing with existing suppliers like NVidia. You don't have to deploy at scale for that to work, but you do have to look like you could if pushed hard enough.
> NVidia has enormous operating margins, so a competitive solution doesn't have to match or beat NVidia's scale efficiencies; it just has to beat delivered cost.
But then that means you have no actual moat against the behemot, right? Your competitor can move into the market as soon as they want to, at much better cost (so at slightly better price)... and Nvidia certainly can adapt much faster around hard hardware specs innovation than a new entrant ever could.
Perl had CPAN already in 1993; Linux and common BSD distros have basically always had package repositories.
"People were responsible for the dependencies they were importing" - I don't know about that. Plenty of us were irresponsible 30 years ago.
reply