LessWrong 2.0 Reader

If you have 3 days worth of storage, even if you completely discharge it in 3 days and completely charge it in the next 3 days, you would only go through about 60 cycles per year. In reality, you might get 10 full cycles per year. With interest rates and per year depreciation, typically you would only look out around 10 years, so you might get ~100 discounted full cycles. That's why it makes more sense to calculate it based on capital cost as I have done above. If you're interested in digging deeper, you can get free off grid modeling software, such as the original version of HOMER (new versions you have to pay).

Even now at $1000/kW-hr retail it's almost cost-effective here to buy batteries to time-shift energy from solar generation to time of consumption. At $700/kW-hr it would definitely be cost-effective to do daily load-shifting with the grid as a backup only for heavily cloudy days.

Please write out the calculation. 

Have there been some recent advances in compressed air energy storage? The information I read 2-3 years ago did not look promising at any scale.

Aboveground compressed air energy storage (tanks) is a little cheaper than chemical batteries. But belowground large compressed air energy storage is much cheaper for days of storage, with estimates around $1 to $10 per kilowatt hour. Current large installations are in particularly favorable geology, but we already store huge amounts of natural gas seasonally in saline aquifers. So we can basically do the same thing with compressed air, though the cycling needs to be more frequent.

(We switched back to shipping Calibri above Gill Sans Nova pending a fix for the horrible rendering on Windows, so if Ubuntu has Calibri, it'll have reverted back to the previous font.)

From footnote 2 to The state of AI today:

GPT-2 cost an estimated $43,000 to train in 2019; today it is possible to train a 124M parameter GPT-2 for $20 in 90 minutes.

Isn't $43,000 the estimate for the 1.5B replication of GPT-2 rather than for the 124M? If so, this phrasing is somewhat misleading. We only need $250 even for the 1.5B version, but still.

wait I just used inspect element, and the font only looks bigger so nevermind

The footnote font on the side of comments is bigger than the font in the comments. Presumably this is unintentional. ^[1]

The paper you describe in your comment would have a lot of it's details filled in by default by the capabilities people inside an AI lab, and the alignment team would outsource most of the details to the people who would want to make the AI go fast.

While I don't think it would ensure humanity's success by any means, I do think that the alignment field could mostly declare victory and stop working if we knew there were no problems that were resistant to iterative correction, since other people will solve it for us.

We're trying to! I normally publish next week's at the end of the meetup, but this week I was sick (and Aella was out of town). Next week's is up [? · GW], I've drafted the week after's and picked the readings for the week after that, so we should be on-time for the next few.

From chapter The state of AI today:

Later this year, the first 100,000 GPU cluster will go online

It's not the first, there's xAI cluster from September, and likely a Microsoft cluster from May.

Even the cited The Information article says about the Meta cluster in question that

The previously unreported cluster, which could be fully completed by October or November, comes as two other companies have touted their own.

Well, it makes things better. But it doesn't assure humanity's success by any means. Basically I agree but will just redirect you back to my analogy about why the paper "How to solve nuclear reactor design" is strange.

I don't think Joe is proposing we find an AI design that is impossible to abuse even by malicious humans. The point so far seems to be making sure your own AI is not going to do some specific bad stuff.

If you solve the latter, you have not solved the former at all; if you solve the former, someone will solve the latter.

Insofar as this is true in your extended analogy, I think that's just a reflection of "completely proliferation-proof reactor" being a bad thing to just assume you can solve.