• 14 Posts
  • 207 Comments
Joined 1 year ago
cake
Cake day: October 23rd, 2024

help-circle




  • For my language, J, I can’t get autocomplete.

    Even though J is a functional language (on extreme end), it also supports fortran/verbose python style, which LLMs will write. I don’t have the problem of understanding the code it generates, and it provides useful boilerplate, with perhaps too many intermediate variables, but with the advantage that it tends to be more readable.

    Instead of code complete, I get to use the generation to copy and paste into shorter performant tacit code. What is bad, is that the models lose all understanding of the code transformation, and don’t understand J’s threading model. The changes I make means it loses all reasoning ability about the code, and to refactor anything later. Excessive comments helps, including using comments as places to fix/generate code sections.

    So, I get the warning about “code you don’t understand” (but that can still happen later with code you write), and comment system helps. The other thing he got wrong is “prompt complexity/coaxing”. It is actually never necessary to add “You are a senior software…”. Doing so only changes the explanation level for any modern model, and opencode type tools don’t or separate off the explanation section.

    LLM’s still have extreme flaws, but article didn’t resonate on the big ones, for me.


  • This is without grid connection. If grid connected, and wholesale exported electricity revenue of 15c/kwh on cold sunny winter day or summer heat wave (72 days potential) would be 3.7 times more profitable than H2 electrolysis (assuming only $2/kg revenue), and generate over $1000/kw. Recent winter storm in PJM area spiked to $3/kwh. It wasn’t necessarily sunny during spike, but if there is a wholesale participation contract, paying employee BEV owners a huge premium is still way more than 15c/kwh. There were very cold sunny places with $1/kwh after the storm. Can consider closing the datacenter those days, but the large battery, can also monetize peak scarcity during the storm.




  • this week‑long autonomous browser experiment consumed in the order of 10-20 trillion tokens

    at $60/m, that is $600M to $1.2B in full price cost, but 1/4 this is current standard pricing. Still, even if a buggy piece of shit, a 1-3m line code project in a week is impressive. OTOH, netscape 1.0 cost $4M to develop, with the advantage of working (though other advantage that it was your web page’s fault for not working).

    They set a very challenging experiment. There is a reason for chromium being a popular base for a browser. The more interesting experiment result is if it is ever usable. Are the bugs solvable by AI.








  • It doesn’t generate CO2 because it is a closed loop, although a leak event is possible. Technically, (if no accident ever happens) then it is small scale sequestration (uses blow up tent). Technology is better than compressed air because it is stored as a liquid at room temperature, and has much better power density on turbine. Commercial viability is greatly enhanced by having a waste heat source, but it’s already better than CAES economically, even with the bubble storage.




  • Adaptation was not defined. It could mean have the world’s rich accomodate the world’s poor, but our failure point has been poltiically that we could not sacrifice just the rich oil company’s extortionist profit power to reduce our energy costs in addition to future sustainability expenses/compromises. How can that broader failure of direct immediate social benefit translate into future political support for social sacrifice for humanist purposes.

    A side note, the only policy that was ever going to, or will ever, stop global warming is a carbon/GHG tax with the proceeds paid to citizens/residents as a dividend. Demonic warmongering for losing wars, along with the divisiveness it invites for inflation and attacks against the economy, with more oligarchist and zionist supremacism as the winning solution is necessary to make you more miserable, and in your misery, lose all concern for human sustainability while consumed with anxiety over your near term future.

    Back to adaptation, perhaps a politically relevant class is relatively wealthy residents of Phoenix. Should “we” give them free homes somewhere that is not an abomination to humanity? Compared to trucking desalinated water into Phoenix for 200 years, digging them a canal from the Great lakes would be a bargain. Florida coastal home owners vote in what is sometimes a swing state… should we build them a sea wall for entire state?

    Surely, with electric bills set to soar 300% because we politically require all the AI datacenters to bring skynet surveilance state that ensures skynet support (or China wins), cannot tolerate an extra 10% electricity increase to subsidize the poor’s access to AC, since Skynet will make the poor unneeded anyway.

    The problem with supporting adaptation, is that you give the politically relevant classes a lifeline to support continued climate terrorism. The lifeline being that they will get bailed out after we are finishing arguing about who deserves to survive/be bailed out.

    The other issue with adaptation, is that it financially costs less to mitigate worsening than it does spending on ever increasing adaptation. Even if you win political support for the stupid option, it is still smarter/cheaper to mitigate instead.