I mean, I think we all recognize that these are gambling sites trying to skirt gambling regulations, so all of their arguments are going to seem ridiculous. “We’re a prediction market, and individuals with specialized knowledge improve our accuracy.” “We allow people to hedge against adverse events, like Elon Musk tweeting over 300 times this week.” “These are financial contracts, not wagers.”

The interviews I’ve seen with “prediction market” CEOs, they’re openly begging for people to trade on confidential, privileged, or classified information, because that’s the source of their markets’ supposed predictive power.

Most of us lived through 2020: saw cities shut down; hospitals using refrigerated trucks as morgues; literally millions of dead. Most of us then saw the vaccine roll out and all of that just went away.

Don’t even get me started on the 1990s. Every new processor generation actually felt faster. Web pages had blinking banners because the creator thought it looked cool, not to advertise a personal information vacuum. There was no better introduction to the public’s absolutely awful sense of style. But I went from talking to international friends for $0.50/minute to free, and it was amazing.

Those companies have extremely well developed propaganda machines. They have to sell their technology and products as benefits to governments (i.e. society) and solutions to chaos (i.e. crime and terrorism), and they have extremely well refined language to describe themselves in positive term. If you don’t look past the company line, it’s easy to believe that the skeptics and warnings are all just FUD from haters, especially when the propaganda pays your mortgage.

Then Palantir goes and publishes an actual fascist manifesto…

To me, that’s the ‘fancy search engine’ mode of AI where it works well and basically focuses the human effort. A needle-in-haystack problem. It might still be missing things, but they’re things you’ve already missed yourself, so no loss.

It’s different from asking Claude, for example, to create a new guest VLAN with limited internet access and access to only a specific service on the private network. For that, you have to 1) trust Claude because you lack the expertise to review, 2) spend time learning the config system well enough to review, or 3) already know the system well enough to check it. 1) just sounds bad. 2) sounds like Claude isn’t saving much time, but maybe helps focus the human where to study, and 3) seems like the human might have been able to just do the job in similar or less time than writing the prompt + reviewing the result.

I feel like the big mistake they continue to propagate is failing to distinguish among the uses of AI.

A lot of hype seems to be the generative uses, where AI creates code, images, text, or whatever, or the agentic uses where it supposedly automates some process. Safe uses in that way should involve human review and approval, and if the human spends as much time reviewing as they would creating it in the first place, then there’s a productivity loss.

All the positive cases I’ve heard of use AI like a fancy search engine - look for specific issues in a large code base, look for internal consistency in large document or document sets. That form lets the human shift from reading hundreds or thousands of pages to reading whatever snippets the AI returns. Even if that’s a lot of false positives, it’s still a big savings over full review. And as long as the AI’s false-negative rate is better than the human, it’s a net improvement in review.

And, of course, there’s the possibility that AI facilitated review allows companies to do review of documents that they would otherwise have ignored as intractable, which would also show up as reduced productivity.

Pretty sure that reflects its stage in the legislative process, not support/opposition. i.e.: out of 100 bills that get introduced, only 1 becomes law.

Because US businesses will only compete and innovate if you force them. Leave them safe behind ramparts of protective trade policies, and they’ll keep coasting on 1990s technology, as the country as a whole slowly becomes a backwater.

fud: Fear, Uncertainty and Doubt. A tactic for denigrating a thing, usually by implication of hypothetical or exaggerated harms, often in vague language that is either tautological or not falsifiable.

That’s just systemd adding a birthdate field to their userdb. Doesn’t require that it be filled out or accurate, and especially doesn’t require it to be validated against a government database. I don’t see it as fundamentally any different from adding a userdb field for favorite color, phone number, or blood type.

Without 3rd party validation, I really don’t see the privacy issue with an age field. Without verification, it is, at worst, one more byte available to hash into a unique identifier, but you can feed that field from /dev/random at every query and poison even that hypothetical.

In the game, you have to improve your properties to charge more rent. In reality, the monopoly can reduce quality and raise price at the same time.

My nephew has a new baby. Her parents are constantly waving their phones in her face; sending pics back and forth; generally doing ‘millenial things’ with their phones when not actively attending the baby. Then proceed to get all freaked out when the baby expresses interest or curiosity in the phone.

Kids mimic their parents behavior and interests. If you want kids with healthy internet use, you have to have parents model healthy internet use.

The trouble with living in a panopticon is that becomes suspicious to not be on a list.

I have a n ESP32 with a thermocouple stuffed down my (gas) oven chimney, so I can tell what temperature it actually is (about 40°F/20°C cooler than the dial).

I have one plugged into an addressable LED matrix, which has yet to get mounted, but will eventually be a closet/dressing light. There’s a few places where I’d like a ‘normal’ warm white light, with the option to switch to a blinding daylight for chores, and maybe a low-light, colorful animated nightlight.

I have a Pi-0w reading temp/humidity/CO2 in a grow tent that’s a good candidate for ESP32-ification. I have an air quality sensor plugged directly into a Home Assistant server that could go on ESP32 if I wanted it in a different location. Humidity in the bathroom, with a controller for the bathroom fan is another good candidate.

If I can come up with a good way to put them on battery, with a 6-12 month lifetime, then temperature in the attic, and on the input/output sides of the HVAC would be useful.

I only one I know about https://socprime.com/blog/cve-2025-27840-vulnerability-in-esp32-bluetooth-chips/ which is a bluetooth thing, presumably meaning that you’d have to be in bluetooth range to exploit it.

My paranoid concern is that I’m going to buy these $2 ESP32 boards from some unknowable Chinese company, and how could I know if there’s an extra, malicious supervisor element added. So, my ESP32 devices live in the ‘untrusted’ VLAN. They could, theoretically, discover each other and send their sensor data to some nefarious broker, but they don’t have microphones or cameras. I don’t even see how they could get enough information to discover my physical address, without cooperation from my ISP.

I was really intimidated by ESP32. Liked RPi, back in the 3b days, because I could comfortably sit in the python interpreter, play with sensor interfaces, and get immediate feedback of what & where I screwed up. Familiarity led me to RPi4 for libreelec and 0w for more sensors.

Recently took the plunge on some ESP32s, though, and, just…wow. I mean, I’m going through esphome, but every sensor and control I’ve checked is just a couple of lines of YAML away. And low enough power that I’m starting to think about batteries. ESP32 is still pretty intimidating for noobs, but the ecosystem that’s grown up around it is fantastic once you get over that hump.

The problem isn’t necessarily corporate services - the problem is corporate services with no practical competition. If there’s an actual marketplace, then enshittification is limited, because you can just hop providers when service degrades. If there’s an actual marketplace, then you can hop providers when some government takes control your provider.

Putting fun services behind the wall of ‘you must be this technically competent to participate’ isn’t going to fix the broken system.

So, these scientists were asked to evaluate a political question, “Is there a link between immigration and welfare support?” using a large survey dataset. Not like they were asked whether temperature data supported anthropogenic climate change. The 158 scientists were in 71 teams and did, collectively, of 1200 statistical tests.

An overwhelming majority of all analyses found no link between immigration policies and support for welfare programs, regardless of investigator ideology. A handful of outlier models, where an effect could be found, show effects that correlated with the team’s politics, but it’s hard for me to look at the mountain of “no effect” conclusions and agree with the statement “politics predicted the results.” “Politics predicted the outliers,” OK.

Actual study: https://www.science.org/doi/10.1126/sciadv.adz7173

Steam can do massive numbers spikes because they have essentially infinite inventory. The whole reason scalping video cards works is that Nvidia can not make as many as people want, even at full retail price. The existence of scalpers implies that Nvidia could raise prices, sell slightly fewer units at higher margin and get greater total return.