The people who would be surprised at this will never see the article in their feeds.

Not strictly true, historically. What we will probably see is the death of is 200-man IT projects. But they always failed anyway and were wasteful garbage so good riddance.

If you set it to "85th percentile of observed traffic" you are selecting 15% to be targets of fines. Why 15 and not 20, or 10?

States with "reasonable and prudent" rather than "explicit speed limits" do a more logically consistent job here. Reasonable and prudent is what we're really looking for - everyone choose a speed that is safe for the conditions of the road, the vehicle, and the surrounding traffic.

The problem is that it's difficult to fine people for that, because it is partly subjective and different for every driver and weather conditions. It's much easier to set an explicit speed limit and then measure speeds. Explicit speed limits exist for the convenience of the courts, with safety of the road users as a distant secondary objective.

If you want to improve safety, then look into "traffic calming" measures. In particular those that cause drivers to perceive higher risk (and research into conditions where drivers falsely perceive lower risk). Even just drawing the lines narrower on a wide street can have an effect. If you design the road right, drivers will naturally choose the right speed for the environment without any need for a road nanny.

Seriously, if a person is betting six figures on the whether an interceptor worked, that person needs to re-evaluate their priorities.

Yes, they can, but they can't do it well. As another example to those in the original link, I asked Google Gemini to compare two PDFs to find differences. The PDFs in question were commuter train schedules with different effective dates. The PDFs had tables with stations and times. Some trains were express trains (skipping stops) and some made all stops. I asked, "The attached are railroad schedules for the same train line during different time periods. Summarize the differences between them" followed by "Are there any differences in the timetables for travel between the X station and Y station?" The output detected that there were no differences between the timetables for these two stations (yay!), but it was *terrible* when it tried to list the actual train times; it couldn't figure out when a train skipped a station. After a few back-and-forths, it gave up on determining the train's departure time, and focused on the arrival time at X station. I gave it hints as to the formatting, and it improved.

The lack of privacy is not new. What is new in the last 20 years is the lack of anonymity that mass surveillance has enabled. I did stupid stuff as a kid, but there was almost zero chance of my stupidity being filmed and uploaded for the world to laugh at.

To clarify, OS X was descended from NeXTSTEP, which was built on Mach and BSD.

"Employee bought company ABC's stock after hearing that his employer is planning to acquire ABC" is already against insider-trading laws, which I imagine is part of the training for Warner (I had to retrain every few years when I worked at a public company). Whether ABC was a former employer is not really relevant. The enforcement of the laws is an open issue in the current administration.

While I agree that I don't understand how this is price fixing, I'm not sure your argument is valid. Standard Oil is a pretty well-known example of producers colluding to keep the price up, but they still kept it low enough that people found a ton of ways to make use of oil from transportation to heating to labor productivity. Using the "loss of demand" measurement we would probably have missed it.

I think the issue here isn't collusion per se, but rather that an information disparity exists and disadvantages tenants and is being perceived as "price fixing" because there really isn't any other mechanism currently to deal with the problem.

One alternative solution would be to level the playing field by finding some way to make tenants and landlords alike have access to the same level and quality of information. I would suggest perhaps all rents and rent offers should be published in a way that anyone can apply their own algorithm on either side of the negotiation.

I have some friends who like conspiracy theories and some who laugh at them.

But not one believes Jeffrey Epstein committed suicide.

I'm not sure how to process that...

Of course some conspiracy theories are later proved to be true (turns out that our governments were lying to us about Iraq having weapons of mass destruction after all... I was called a conspiracy theorist by these same friends at the time for saying this). History is replete with examples like this. So, I don't know how any study can claim to have the monopoly on common sense here.

Why have automated calls at all? Sports are social events where people got to see other people who have trained to peak human performance compete against each other at popular games.

I don't want to see perfect play by optimized automatons, I want to see the earnest best effort on the part of the participants and argue about different plays with my friends later. The players will make mistakes, and so will the officials. That should just be part of the game.

It's a UI prototyping tool. My company's UI folks seem to like the ease of developing prototypes with it, but its organizational system is nonexistent. I can never find any files; the only way I can access any of the files is via links people send me. Maybe this is due to the way the permissions are locked down at my company.

Only if you don't live near a Staples, CVS, Walgreens, etc. You'll get even better results because (at least CVS but probably everyone) has dye sublimation printers, which are vastly better for photographs than color printers.

For other color prints, Color laser printers are not that expensive, but I haven't had much success. I'd recommend stick with B&W, and outsource color stuff to office supply / pharmacy type stores.

B&W laser is so vastly cheaper and lower maintenance, particularly in a sporadic use case, that it's not even a competition - Last time I bought a laser printer, the starter toner was good for 1500 pages at 5% coverage, and it worked fine even if I last printed something 8 months ago. Compare to a typical inkjet in which a full toner pack is good for 200 pages at 5% coverage, has to run ink through every so often to keep the tubes clear, and (for the last HP inkjet I owned) prints black text using all colors despite having a full black cartridge.

Caveat: if you are planning on printing addresses on envelopes to be mailed as a substantial use case, go with the inkjet. Toner fuses to the outside of the paper, so it can get stripped away by the sorting machines.

At least make it sound as if you're not being paid to write this, for crying out loud.

Back to the article: basing any insight on the number of job adverts is bogus. Everybody knows that most adverts are for jobs that don't exist (ask any recruitment agent that you know socially).

And the skills available come from self-reporting. Guess what? In a boom, everybody says they have the skills. A friend of mine is selling herself as an AI expert when she graduated in fashion and has worked in retail for decades. Don't get me wrong. She's a smart cookie in her domain. But AI expert? WTF?

Do you think when I said "script" that I was literally referring to a kind of "Protocols of the Elders of Zion" document rather than employing a simple metaphor?

Seriously?