aoanevdus

This thread is cluing me in that I need new wording to describe my rare consumption of alcohol.

It’s tempting to anthropomorphize a company like Google, and assume that every behavior is part of some evil master plan. Just as often, it’s some small group of people within the company making a dumb mistake. I’d guess there is some team tasked with reducing the “app spam” problem, where there the store and (app review process) is crammed with thousands of near-identical apps, torturing the naive user with ads as they attempt to perform basic functions.

This targeting of this warning is over-broad, preventing honest new app developers from getting traction. That’s bad for the long-term health of Android’s app ecosystem, and a competitive disadvantage against iOS. There’s probably some other team at Google who is responsible for improving the development experience for Android, who hates this new warning.

Talking about the harmful outcomes of this warning, it’s good to get the news far and wide and try to get it fixed.

Analyzing why the thing got pushed in the first place, it seems to me a symptom of the challenge of coherently managing a hundred thousand employees.

The common pattern of requiring three days a week of in-office time makes it much harder.

When a large institution is faced with uncertainty about the future, it’s both feasible and prudent to make plans that account for multiple future outcomes. In this case, it makes sense to do both of the following:

1) Fight the administration in the legal system.

2) Plan for the case where some of those legal fights are lost.

Supposedly there are hundreds in Concord.

In pure math at a school like Harvard, the standout kids like the ones in that quote are probably trying to become tenured math professors. There are very few such positions available. You can shoot for the stars, and if you succeed, make about the same as the average software engineer. More likely, get stuck a postdoc. So most students give up pure math at some point. If you realized you weren’t cut out for it in freshman year, you got a head start over the people who got a math phd before finding out the hard way.

This pressure didn’t exist in computer science because there were plenty of tech jobs for anyone competent (not sure if that’s still true in 2025). And you didn’t need to be a genius to build something cool.

When I was a kid, an adult told me that I should stop using “basically” as a filler word because people will interpret it as an insult to their intelligence (ie. “You’re not smart enough for the whole thing, so I will just tell you the basic version”). I’ve been attentive to the way other people use the word ever since, and I think they have a point. Some people say it very frequently and don’t mean anything by it. But a good chunk of the time, it does seem like there is a status game going on when people use that word.

Quite a tangent, but what if we apply this logic of informed consent to property? If a person without a will dies, should we leave their house abandoned until it decomposes? Automatic organ donation is like probate for bodies.

I thought that was because the lab meat competes with farmers. I can’t think of any similarly situated rival to the lab bodies for organ transplants that would lobby against it.

Isn’t trading off of delayed disclosures basically being dumb money? You’re in a similar position to a pump and dump victim.

Would they need a veto-proof majority for that to work? You’d need like half the republicans to go along with that.

Don’t assume that because the article depicts this competition between the US and China, that the authors actually want China to fail. Consider the authors and the audience.

The work is written by western AI safety proponents, who often need to argue with important people who say we need to accelerate AI to “win against China” and don’t want us to be slowed down by worrying about safety.

From that perspective, there is value in exploring the scenario: ok, if we accept that we need to compete with China, what would that look like? Is accelerating always the right move? The article, by telling a narrative where slowing down to be careful with alignment helps the US win, tries to convince that crowd to care about alignment.

Perhaps, people in China can make the same case about how alignment will help China win against US.

Why even go through the effort of making a criticism when you can satisfy the urge simply by pointing and implying one exists?

https://apnews.com/article/russia-poland-germany-sabotage-ca...

November 5, 2024

WARSAW, Poland (AP) — Western security officials suspect Russian intelligence was behind a plot to put incendiary devices in packages on cargo planes headed to North America, including one that caught fire at a courier hub in Germany and another that ignited in a warehouse in England.

Poland said last month that it has arrested four people suspected to be linked to a foreign intelligence operation that carried out sabotage and is searching for two others. Lithuania’s prosecutor general Nida Grunskiene said Tuesday there were an unspecified number of people detained in several countries, offering no elaboration.

People have been sending explodey batteries by air freight. In that context, requiring batteries on a plane to be in the cabin where they can be located, accompanied by the owner of the battery could be a good deterrent.

IMO this is a use-case where the added latency of cell phones is a downgrade.

At Amazon it’s common to have terrible, week-long oncall shifts with many repetitive pages. At Google they have shifts that follow the sun and they get PTO to compensate for being oncall during the weekend. Both jobs pay similarly. And I think most people joining Amazon don’t know about the oncall they are in for.

I’m curious what would be considered the industry standard for fact checking in tech. Does Google Search, Apple App Store, TikTok, Snapchat, Amazon store, etc. apply fact-checking to the content posted by users/sellers?

Or more abstractly, is fact-checking the responsibility of authors and content editors, or of platforms and infrastructure that spread the content?

Assuming you’re driving a sports car for fun and not to win a race - isn’t feel the whole point?

By this metric I would have got into any school I wanted, but that’s just because I put an exceptional amount of effort into preparing for the test. My grades and extracurriculars weren’t top-notch. I did go to an elite-ish school and it was clear that many other students deserved to be there more than me (ie. were able to contribute to society more in various ways), and in my view that difference was legible in the admissions process.

Neat idea. I think I’d like it to maintain a global history and local, with some UI affordance to remind me both exist and perhaps what are the last few entries.

A lot of my commands would work in any directory, but sometimes the directory matters and I run it in the wrong place.

What’s a “mechanical process”? If I read The Lord of the Rings and it teaches me to write Star Wars, is that a mechanical process? My brain is governed by the laws of physics, right?

What if I’m a simulated brain running on a chip? What if I’m just a super-smart human and instead of reading and writing in the conventional way, I work out the LLM math in my head to generate the output?

The structure of the files an app uses internally is undocumented and not intended to be a user-facing API. Who wants to be responsible for handling every insane thing the users decide to do to those files?

There are practical reasons to understand the effects of both cannabis and alcohol. It’s not like we can only care about one of them at a time based on which is winning the social harm olympics.

As a former christian, I don’t agree with that at all. I consider my ability to stop believing to be an opportunity that I was lucky enough to receive based on being exposed to the right ideas at the right time in my life. I wouldn’t for a moment look down on anyone who kept believing because they had a different life experience.

I don’t get it. The labs have regularly made improvements that dramatically lower the cost of training an equal-performing model. When they do this, they also train a larger model with even higher performance. This time, DeepSeek did the first part but didn’t do the second. Now every lab in the world will throw their compute into the effort to replicate and beat DeepSeek’s model with larger scale. It’s not like everyone is just going to say “well I guess AI is smart enough now, no point improving it anymore!” and stop building bigger training clusters.

If anything, r1 makes even more GPU demand likely, since it mitigated or at least delayed the risk AI hit a dead end (in which case, ceasing development may actually make sense).

Would it be different for non-e bikes?

I wonder how many board seats Microsoft controls compared to Altman loyalists.

Edit - last year, Microsoft quit the OpenAI board https://apnews.com/article/microsoft-openai-artificial-intel...

I think Microsoft’s shares don’t have voting power.

I’m sure if I get hurt on my bike, my medical insurance coverage will not care at all whether the injury is from riding an e-bike.

If I hit somebody else on my bike… I don’t think I have insurance for that liability? I don’t think my auto or homeowner’s insurance policy mentions me riding bikes at all, let alone an exclusion for DIY e-bikes.

I doubt Microsoft is dictating terms here. I think the scene is many organizations competing for the opportunity to supply OpenAI’s compute. And OpenAI making choices based on business, technical, and political considerations.

Although in that case, it wasn’t the back of the head, and one of the shots appeared to be non-fatal. So a terrible way to go, but not implausible.