dgacmu

Many personality disorders have a strongly heritable component.

I believe you need to divide by another ten

For me, habit. A habit I'm getting tempted to break by exactly the grumble in this post. This has been an awful ~two weeks for maintaining infrastructure and the ubuntu lag on this specific issue is contributing to the awfulness.

That doesn't math all that well. The coefficient of performance of a heat pump is better when the outdoor temp is higher - did you mean the temp outside is -above- 20F?

You're also assuming time of use billing, I assume.

I don't disagree, btw - gas is cheaper than a heat pump for me also, which makes me sad. My only point was that you have to compare against them and not resistance heating. (That's slightly wrong - if I were to use a heat pump during the shoulder season and my boiler when it's colder it would be cheaper in operating costs but lose on capex)

Because these are bad comparisons, a heat pump water heater and a heat pump house heater use substantially less electricity. Resistance heating sucks and will lose to burning gas locally.

It does, but the video is later than their original profile and not linked, and I thought it was interesting actually seeing him in action. It's one of the higher quality videos of him.

Funny that it's in the new yorker - they had a great video of Apollo Robbins about 13 years ago demonstrating some pickpocketing: https://www.youtube.com/watch?v=LoUSO_Mj1TQ

But we generally don't need to satisfy the peak with battery - peak US demand occurs while the sun is still up, in July, so you can satisfy a large chunk of that with solar still. The hard part is those ~4 hours after solar stops producing enough and it's still hot and people are still up. So what we need is several hours of about half the gap between {nuclear, wind, hydro} and peak. Remembering that we also use even non-pumped hydro as a form of storage, by drawing on it more in the periods just before and just after sunrise/sunset.

For example, if you look at the energy production across PJM from two days ago: https://cdn.bsky.app/img/feed_fullsize/plain/did:plc:hiovtnl...

(Using that b/c I run the bot and it's data I'm very familiar with. You can see similar patterns in CAISO but they've already got so much solar it's a little more messy/duck-curved by all of the behind-the-meter solar.)

You can see hydro coming online from roughly 16:00 - 23:00. There's still a big gas peak that we'd need a lot of storage to shave, and you're right that it's a long way to go before we do that, but it's not quite as bad as "we need enough battery to satisfy 100% peak demand".

Not with the disrespect they've shown to Arsenic, Astatine, Barium, ...,

They finally gave the $20/month pro tier access to argon in antigravity so I tried it a few days ago on one small thing and ...

it actually did really well.

But it still annoys me: I told it to _tell_ me something, and instead it _implemented_ the something. The previous models were way too eager to just jump in and do and this one seems no different. But the optimization it identified was both really solid and just a good engineering choice, and better than what Opus had been suggesting.

It also gave pretty good feedback on a bit of writing.

I agree with the threat to the expensive providers but why is openrouter use representative of overall token spend when you'd think many companies are buying directly from anthropic or openai?

You're mis-reading my post. _Pittsburgh_'s air isn't great. The air inside my house is great, thanks to a lot of air filters.

The greater Pittsburgh region is #11 on the most polluted cities in the US by short-term particle pollution: https://www.lung.org/research/sota/city-rankings/most-pollut...

It's not Mumbai. But it's bad by US standards.

The thing that's not fully optimized are the CO2 levels in my house. As I've continued improving the air sealing in the house to reduce my heating and AC costs, it's something I want to address a little more in the future through active ventilation.

Depending on the month, my kitchen AQI averages between 1-4, representing PM2.5 values of 0.1 - 0.8 ug/m^3. Bedrooms are likely better as they don't have the gas stove.

CO2 while occupied is generally under 750; I'd prefer to keep it under 700 but we also balance that with energy efficiency for that make-up air; I don't have an ERV, so we suffer a little there.

(We don't live by a freeway but - (a) Pittsburgh's air isn't great; and (b) the canadian wildfires were pretty bad here at times this summer, and that was what caused that 0.8 ug/m^3 pm2.5 spike).

We have, um, a lot of air filters left over from covid, though, so we're a little spoiled in terms of being able to manage AQI easily.

The actual number needs to be interpreted carefully but the change is certainly real.

Because Google doesn't have much of a financial interest in that kind of paper. Drug companies have a financial interest in studies that evaluate their drugs.

If you want a CS analogy, the google-co-authored papers about internet policy do get a lot more scrutiny. And rightfully so! That doesn't make them bad -- many of these are written by friends of mine whom I hold in great esteem -- it means we have to be aware of the potential for all of us to be influenced by the money.

Dude. From the paper itself:

"Competing interests G.A.M. serves as a research consultant for Merck, GlaxoSmithKline/ViiV, and Gilead. M.J.C. serves as a scientific advisor for TruDiagnostic. V.D. and R.S. are employees of TruDiagnostic. All other authors declare no competing interests."

These are known risk factors. I've never said drug companies are bad. I literally take corporate funding for my own research. But you also don't get to stick your head in the sand and ignore the competing interests that all make the research more likely to show favorable results. Again for those in the back: These results probably _are_ correct! But do they generalize? Will they hold up on replication with a larger population? Do they matter, or is the proxy they're measuring not actually relevant for long term things in humans? Dunno. That's the thing about a _statistical likelihood of bias_ -- you have no idea exactly how to apply it to a single paper other than to say "ok, known possibility of bias, interpret accordingly."

In addition to the funding bias wikipedia article someone else already cited:

https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.MR...

> Sponsorship of drug and device studies by the manufacturing company leads to more favorable efficacy results and conclusions than sponsorship by other sources. Our analyses suggest the existence of an industry bias that cannot be explained by standard 'Risk of bias' assessments.

And no! It's not unusual, nor does it mean the results are wrong. It just means there's a known bias towards results that favor the sponsor's interests that needs to be taken into account when evaluating the research. Again, that doesn't make it wrong. But it should make you cautious before thinking that a single study (in a very specific sub-population, at that) is necessarily going to generalize. It might. I hope it does! GLP-1's are awesome. But cautiously.

It's not but there's a lot of historical evidence that those situations are more likely to produce research that looks favorible to the sponsoring company. So it means you have to be a little more careful when looking at those results.

Er, yes, that's exactly the behavior you would expect of something growing exponentially - it looks slow at first and then suddenly overtakes things. 23-27% annual growth takes a while to really get moving.

I decided to try something a little weird research wise the other day analyzing about a thousand older research papers and burned through my $100 chatgpt sub + 3 resets in three hours and only finished half of them. It's the first time I've done that, though.

They've been mapping, or whatever they do, Pittsburgh for almost a year now, and it was pretty funny watching the cars navigate around massive snow piles last winter. Curious to see how the rollout goes here. <Insert popcorn emoji.>

For thousands of years, we used time synchronized to the sun to some approximation that was off by minutes, not microseconds, and nobody cared if the current time relative to the sun was off by a second because they had no way of actually determining that. And then some jerk came along and invented computers and precision timekeeping and suddenly things depended on things like time not appearing to move backwards by as much as a second. :)

But that's a very modern perspective:

1950: Lighting was 30% of total electrical demand [1] 1993-1997: 112 TWh, 9% of demand [2] 2020: 81 TWh, 6% of demand [3]

Couldn't find total used in 1950 but the point remains: Lighting used to be a major source of consumption and now it's pretty minor, and at least the trends in the last 30 years have been driven by efficiency gains, particularly following the energy crisis in the 70s.

[1] https://www.researchgate.net/publication/222531686_Exergy_po...

[2] https://eta-publications.lbl.gov/sites/default/files/lbnl-40...

[3] https://www.eia.gov/tools/faqs/faq.php?id=99&t=3

As of now Argon is available in the android app, at least.

I'm not the person to whom you're replying, but I just tried one of mine:

"I'm interested in understanding the rate of VOC dissipation from fresh oil paint on a surface versus the ventilation rate in the room in ACH. I would prefer solid academic studies of this, followed by analytical evaluation, followed by references to non-quantitative sources. Tell me what's known about this."

that I had run in ChatGPT earlier this week. (Yes, we just had our bedroom repainted, and it's TERRIBLE and I hate it, why do you ask? :-).

I didn't try it first in Gemini Flash, but I did try it in whatever model is used for Gemini Live and the answers were terrible. The ChatGPT answer (I think on medium) was pretty good but it threw in extraneous formulae that weren't actually germane to the analysis. The Argon 4 query I just ran was honestly the best of the answers I've gotten -- it was clear and really well-structured from the perspective of "here are the three big phases that researchers break down paint drying into and what the differing effects of ACH are on those phases".

I'm totally going to give this one a try, after basically giving up on 3.8 flash for most things despite having a free $20 pro subscription (from buying a phone).

Edited to add:

Argon4 transcript: https://share.gemini.google/18Gr5liH1YWR

ChatGPT transcript: https://chatgpt.com/share/6acedc72-ff18-83ea-8909-104cdbc6ac...

No, not in this case. The assumption of unit cost arithmetic is simply not physically realizable. It's probably easier to understand the integer case: numbers take bits to store and represent. In a typical computer, you would store numbers as either 32 or 64 bits, where 32 bits could store a number up to about 4 billion (or a signed number between -2 and +2 billion). If you want to store and multiply numbers larger than that, they take up more space and they require more work on the part of the processor, and fundamentally, more energy - adding two 128 bit numbers takes twice the work as adding two 64 bit numbers.

The same applies to the precision of floating point numbers. What Ben is arguing is happening in this case is the AI has found a way to hide work by moving it into operations that, in one (unrealistic) model of computation, are assumed to cost the same regardless of how large or precise the numbers are. Other models of computation have a more accurate model of the cost of these things, but that was not the model the AI chose to apply in this case.

I would in general agree with you, but having read through the journal paper: I think the fraction of general audience newspaper readers who could actually read that paper is very low. I come from a science background, spent time dabbling in time and frequency measurement enough to understand some of the concepts, and still found it a painful slog. (I am very much not a physicist!)

I still think it's the right practice for newspapers to always link to the primary source, but in this particular case I understand why they might make an editorial decision not to.

The actual journal article, since NYT didn't bother linking to it: https://www.nature.com/articles/s41586-026-11084-4