msandford

100M active time series is good information, but what's the data rate for each time series it can handle? One update per minute or 10 per second? There's a factor of 600 difference there. Neither is obviously insanely the wrong update rate.

You don't even have to ban fishing entirely, just reduce the fishing to a point where fish populations increase. It might mean that fishing has to be at half the level it is now which still isn't great for the people impacted but you don't have to go 100% outright ban.

That's a great idea, thanks for sharing!

This is the same problem as the "AI drives the car until it can't" situation. Feel free to doze off so long as you can wake up and instantly have world class racecar "save the situation" reflexes. Hope isn't a strategy but that's what all of this feels like.

My feeling about all the corruption talk is that everyone is at different points on the "oh yeah now I see it" curve. I've noticed that I'm really annoyed at people who are just now getting it as I've been seeing it for 20+ years. Because people who are further behind didn't see it before they have no idea how bad previous corruption was and only see the new corruption as a unique evil, unmatched by anything that ever happened in a million billion years. Which then makes for very tiresome arguments about how bad it used to be back in the day. They actually argue that it used to be fine, but NOW it's bad.

Anyhow kind of a rant but it's an insight I had as to why nobody can agree on anything despite lots of folks probably being honest/good faith actors.

You know that the usage has gone up something around 1000x since AI right?

Nobody outside of actual engineering or engineering adjacent like construction used load bearing up until about 6-12mo ago. Now it's everywhere.

Isn't this also more unique since they're trading futures contracts vs whatever a traditional sports bet is? I thought it was kind of a clever end run around the law where taking down their gig would also mean that you have to attack commodities futures contracts as well and that the Congress would be loathe to do that lest they accidentally outlaw the stock market.

When I say "clever" I don't mean"good" I mean "crafty". You acknowledge someone has game without condoning their behavior or extolling it as virtuous.

It was literally Obama who did the things mentioned that had bipartisan support.

https://en.wikipedia.org/wiki/Killing_of_Abdulrahman_al-Awla...

I suspected that this was the case as the president can't just invent new laws on a whim. They have a fair amount of leeway on their interpretation of laws re: directions to the administration on how/when to do things but they can't just invent a law from thin air.

Also I mean what constitutes a generator if it isn't something connected to the grid? If my car has an engine and it also has a small inverter and an AC outlet is that now a generator and subject to generator laws in addition to car laws?

All we have to do is improve it by about 100000000000000000000000000000000000000000000000000000x and it might be able to power a phone for 1 second

I'm confused about the idea that the project was a net negative. Based on what? River cleanups are happening that weren't happening before. That seems relatively positive.

Is the negative that they spent too much money on the trawling that didn't work?

The title of the article is kind of technically correct but still misleading.

It is true that The Ocean Cleanup "we're going to trawl the ocean for plastic" approach isn't working terribly well.

But they did pivot to stopping trash from washing out of rivers and into the ocean which is working. Stopping the mess from growing is also a very good thing.

This seems like some combination of the enemy of good is perfect and also sour grapes that someone else is doing the thing.

What iron mountain does know is which drives are owned by oss. What they don't know is which drives owned by oss have the station's data on them. Hopefully nothing is comingled and drives are dedicated to a customer. Seemingly only oss knows and they went bankrupt so who knows what shape the records are in.

It would be nice to see change in debt numbers vs change in deficit numbers. I think only Clinton ran a surplus in the last 30 years.

Yes it's true we shouldn't let perfect be the enemy of good. But good also has to be good enough to solve the problem without creating more problems faster than it solves them.

It's fine to propose a solution but if people aren't convinced just throwing an aphorism grenade at them isn't productive.

Think about gasoline consumption in an economy. If gas is $1/gal in today's dollars you can make many trips, they're almost free. At $3/gal people do some prudent economizing. At $10/gal people stop driving and take the bus, carpool, etc.

Making the price go up 10x won't cut consumption by 90% though, maybe only 50%. Trades still need to get to the job, food still has to get delivered to stores and some people who make huge salaries will still drive to work.

During the pandemic fuel demand didn't go down by 50% and yet the price dropped dramatically. Sure office workers stayed home but that's not 80% of the jobs. Think about how much the gas price dropped vs how much traffic was still on the streets. It almost certainly wasn't a linear relationship.

If you're a market maker? It's the whole job. They're Federally regulated to do precisely that.

If I'm poor I'm still indirectly exposed through the stock market. Employers stock goes down I get fired. Employer has better robotics because of AI? I get fired. Datacenter wants to get built where land is cheap? I get evicted.

Just because I'm too poor to own stocks doesn't mean stock prices don't affect me. That's indirect exposure not direct exposure.

How do I get zero direct and indirect stock market exposure, no electricity price impact, no RAM or GPU price impact, etc?

I'd love to live in a world where AI firms bidding these things up doesn't affect me but I'm really struggling to understand how they aren't impacting the market.

The article concedes that there is an actual base load that's real (so not a myth) but then talks about another sort of related thing and says that's dumb.

Okay, but that isn't really base load. That's the other thing. Base load is real. It's statistics.

Nobody is safe from a nation-state "attack" they'll just go threaten your providers to give up your data. Passwords written on paper are probably safer than a centralized password manager for almost every circumstance other than a government coming after you.

I recognize the "esp32 perched on top of a circuit board" from the picture. I've got a couple of those at my house right now, I'm on the second version of one board and the third of another. I think I'm going down a very similar road for a different product. Would you be willing to some paid advising? I'm an EE so I don't need hand-holding on the actual boards or programming, really just looking for more information on everything else around the edges. Either way, thanks for the article. It was nice to get some validation of my development process.

Maybe you didn't read that it was the history of how Britain eradicated rabies?

For most of history until the rabies vaccine if you got bit by a dog there was a 5% chance it killed you unless you amputated the affected limb.

These are very high stakes for a dog bite.

1. They do a really bad job because you can't get fired

2. My package is delivered late yet again

3. I don't bother calling because the other 8 times I did I wasted an hour, nobody did anything (because nobody can get fired) and I didn't get any money back

4. Look how great we're doing! Complaints are down

5. Give us more funding because we're doing such a good job

This is just as likely of a spiral if we keep it publicly owned. It's also not good. How do you get good outcomes with.no accountability?

My problem rate with Fedex or UPS is maybe 0.1% of packages. I can't remember the last time I had a delivery issue.

Just this week I had a package that was supposed to be delivered by Monday that lost tracking and didn't show up until Wednesday.

It might be "basically fine and good enough" but it's definitely not "amazing and completely beyond reproach" at least in my opinion.

Prior to COVID I seem to remember that the point of a vaccine was so that you didn't get the disease at all. That's where the herd immunity argument comes from. If 85-90% of the herd has immunity (immunity meaning you can't catch it at all) then even the people who can't get the vaccine are protected because the susceptible are too spread out for the infection to spread.

Do you see how this absolutely requires not just a lowered infection rate with lowered severity of infection? You need a less than ~10% chance of infection in the vaccinated/otherwise immune for the math to work. If the chances of infection after vaccination are still 80% then there is still a large reservoir of potential carriers and you don't have herd immunity.

For most of history what people expected out of a vaccine was immunity from infection not still getting infected but with less severity.

Huh that's interesting! I found celery to mostly match my expectations. I used it in a couple of django apps. My only real foot gun was around having to set an EAGER setting for local development or tasks never got executed.

How did you find your expectations and celery's actual semantics to be different? I'm trying to document well and it seems like I might have some implicit assumptions that I could make explicit, but I don't know what they are since they're already in my head and matching celery it seems.

This really seems like a solution in search of a problem. Maybe it would be useful for pro tour riders, but I would guess it'll be banned as not in the spirit of cycle racing. Recumbents would really change the game in the pro tour for speed but the UCI doesn't allow them for the same reason.

The idea that's been floating around in my head for the last few years is something like "it's being trained by the data produced by people, it's going to have many human flaws as a result"

I always thought it had to do with which way the vessel leaned while turning. Boats will lean into the turn, ships out. I guess that really only tells you if that's a displacement hull or planing hull, though.