I’m too lazy to look for any of their documentation about this, but it would be pretty bold to believe privacy or processing claims from OpenAI or similar AI orgs, given their history flouting copyright.

Silicon valley more generally just breaks laws and regulations to “disrupt”. Why wouldn’t an org like OpenAI at least leave a backdoor for themselves to process API requests down the road as a policy change? Not that they would need to, but it’s not uncommon for a co to leave an escape hatch in their policies.

I gotta say, the C02 number seems very high to me too, just got that from a quick search and saw that a couple of times. I haven’t investigated it closely tbh.

I wasn’t aware of the mining differences between uranium and thorium, that is encouraging.

Regarding the waste, that’s a fair point as well. Thanks for the response! Interesting points.

I used to be very pro nuclear energy. Besides the waste and the occasional meltdown it seemed like a no brainer as a renewable supplement. After learning a little more about it though it just seems like we have more runway for positive growth with wind and solar than nuclear, but I’d be happy to be proven wrong.

From what I understand nuclear in general is (at least now) a dead end as a climate change solution.

1. From planning time to turning on the reactor is something like 15 - 20 years (note, that’s longer than the global average of 7 years for construction, because construction is not the whole picture)
2. It’s difficult to have more than 1 plant project ongoing simultaneously due to the scale and complexity
3. Nuclear plants take a lot of C02 to construct and maintain. The fuel has to be mined, resulting in emissions, and the amount of concrete required massive. 1 ton of concrete creates .8-.9 tons of C02, and a nuclear power plant has hundreds of thousands of tons of concrete in it.
4. We still don’t have a good answer for handling nuclear waste.

Maybe at some point in the past nuclear could have resolved many climate change issues, but between project time, initial emission cost, and waste, it just doesn’t seem viable anymore.

Gotcha, all that makes sense. 6 years seems like a good sign. Thanks for the info, definitely checking the org out.

Are you a member? How is it?

Are they? I watch YouTube on Firefox all the time, seems fine on my machine.

I think maybe 5+ years ago there were some performance issues caused by YT relying on features that were only implemented in Chrome, but I don’t recall having any issues wrt that for years.

Woah, definitely need to check this out. I wanted to slap guix system on an old laptop but had issues with proprietary drivers, very curious to see what workarounds people have had luck with. Otoh I barely touch this computer, and NixOs is running fine on it…

In addition, hardware developers reinvent old ways of doing things and only learn by making all the same mistakes that have been made before. It’s sad, but true.

This same criticism is validly launched at software devs all the time lol.

One thing I’ve anecdotalally seen and heard is hardware guys indicating that something is rock solid and solved because it’s old, so building on top of it isn’t a problem. Obviously we have to build on the old to get to the new, but if we just skip auditing hardware due to age we end up deploying vulnerable hardware globally. Spectre and Meltdown are an interesting example where I’ve heard from at least one distinguished professor that “everyone” believed branch prediction design/algorithms were essentially done. Was it adequately assessed from a security POV? Clearly not, but was it assessed from a security POV in general? I have no idea, but it would be nice as a tech enthusiast and software guy to see the other side of the fence take these things seriously in a more public way, in particular when it comes to assessing old hardware for new attack vectors.