This article strikes a tone that suggests the author believes "big tech" is being "petulant" about end to end encryption, and it seems to be trying pretty hard to conflate other issues, like blocked mergers, with the right to privacy and proper implementation of secure communications.

Signal, mentioned by name in the article, is notably NOT big tech, not even a for-profit company, and they're saying the same thing everyone else is saying: they will exit the market if this regulation goes into effect. That should speak volumes and its stunning that UK politicians have not seemed to notice.

Digital cameras use either "global" or "rolling" virtual shutters. "Global", as described in this article, is just a perfect "all pixels start and stop capturing light at the same time". This is more expensive to make because you need to have extra circuitry and memory on the sensor to read out the value of all sensor pixels more or less simultaneously.

"Rolling" is where the sensor rows are exposed then read out one at a time in sequence, from top to bottom. This is cheaper to make because you only need to be able to read out one row at a time, so fewer A/D converters on board, less memory, etc. This is roughly equivalent to having a "shutter" with a thin horizontal slit that slides from the top to the bottom of the sensor quickly on a film camera. It does what you might imagine to fast-moving objects, distorting them so they are slanted, squashed, or stretched.

There's actually a third secret mode some budget digital cameras use called "global reset release". These are rolling shutter cameras that are setup to start exposing all the rows in the image at once, then read them out as quickly as it can. This means that the rows at the bottom of the image keep collecting light while the top of the image reads out, meaning the image gets brighter and blurrier as you get towards the bottom. However, if you include a mechanical shutter and snap it closed once the desired exposure time has elapsed, the sensor can take its time reading out the rows one-by-one while the shutter protects the still-active sensor from capturing more light. Its a neat idea, but it greatly increases the time between shots so it's not too popular.

Starry is the only real competitor to Comcast's ISP monopoly in my city and, so long as you live in a big building and qualify for their actually good rooftop equipment, their service is fine. They are a little bit pricey and their speeds were just ok, but I'd pay the same and get a little less to stick it to comcast.

Unfortunately, I never got the chance! I was never able to get them to service my building. I own a condo in a mid-size building, I'm on the condo board, and I have the authority to let them build up infrastructure on our roof. We have good line of sight to other Starry-served buildings. I contacted Starry several times to tell them, hey, please call me back, we want to give you money! Install your equipment here! I was basically ignored, save for an automated reply saying that "service was not yet available in my building". I guess a human being didn't read the message. Given that experience, I'm not exactly astonished to see them in bankruptcy.

Hopefully they bounce back with better management, more focus ,and a better sales/install team. Comcast needs a rival and their technology seems solid.

I dunno about fawning, but the inclusion of his name is odd every time OpenAI gets brought up, considering he resigned from the board 4 years ago and doesn't have anything to do with the company's decisions or operations anymore. Microsoft is by far the largest single funder, contributing a billion dollars on their own.

I don't know, everything in the article sounds reasonable. The landowners and those affected by the mining operations should get both a say in mining practices and a large cut of the profits. These are their natural resources and their health risks after all. They ought to negotiate terms that make the resource extraction worthwhile to everyone involved and if companies find that too onerous then the lithium should be left alone.

Theres no problem here, just things working the way things ought to work. Nobody has any obligation to "EVs" as a concept to poison their own water supply or allow submarket value resource extraction. Companies have no natural right to a resource on their own terms just because it is there.

If any of these ads are obligatory to all tiers or require a count to 5 then button press to skip, I will cancel Netflix service in a microsecond and shittalk them every chance I get. Zero tolerance, scorched earth, no ads.

Apple's latest laptops use USB-C charging in addition to magsafe. In addition, the new magsafe cable is USB-C on one end and uses USB PD for negotiation.

Its unlikely USB-C charging will be removed as Apple uses it for docked power in their own products. They've simply added another, better charging port, which would be allowed under any regulation that wants common standards.

These are, without a doubt, some of the lowest quality comments I have ever seen on slashdot, and I've seen stories where the comments are just "GNAA" copypasta and swastika ascii art.

The accusation that the timing of this announcement is political, the utterly confused, anti-scientific information about vaccine efficacy, the basic misunderstanding about disease epidemiology, misunderstanding of cause of death and hospital reporting. Most of it willful ignorance and politically driven. Not one single insightful thing said about the announcement or the trials.

  I am not sure why slashdot comments are sub-facebook, sub-twitter, sub-4chan quality but here we are.

The people who died in autopilot crashes after they decided to to not monitor it and take a nap or watch netflix might disagree.

Like, if they were still alive I mean.

Real answer to glib comment:

This is a purchase made in an app, not an in-app purchase, to be brain-meltingly legalistic. Because the purchase is primarily for the car, and doesn't actually change anything in the app, it falls under the rules for shopping apps. Likewise when people use the tesla app to pay for charging, it falls under the rules for shopping or financial transaction apps (amazon, venmo) where neither apple nor google would take a cut.

It doesn't matter at all if this guy legitimately "butt dialed" the purchase or not. It could be an accident, it could be a deliberate purchase with buyers remorse, or it could be a technical glitch. It could be anything. It doesn't matter. We don't live in a world where the default condition is that all sales are final. If a purchase can be undone, and it wasn't made under clear, non-refundable conditions (of which Tesla has an official refund policy), then the customer is owed a refund if they would like one!

This is a story about a person who wants a refund on a purchase, and seemingly cannot get one because the company in question has extraordinarily poor customer support. That's all. The rest is a circus and apparently there are a lot of clowns here.

*dons clown nose* and honestly, if I had to guess, I'd say this was not a butt dial but a fat-finger by a service center employee. Service centers can order these upgrades on the back end, and it seems like, from what is shown in the customer's app and they way Tesla so confused, that this is what happened. If the upgrade were purchased via the app, it would be refundable there. Service center upgrades send a push notification but cannot be undone by the app. Most likely someone ordered an upgrade at a service center and the service center employee opened the wrong account and clicked the button. It auto charged to the payment method of file for the account (used for super charging, so most accounts have one) and here we are.

Some of this is spillover bias from photography, both the historical chemical process and the modern sensors and algorithms in smartphones. Both are tuned to produce nice exposures of lighter skinned people, while darker skinned people are not given as much consideration. This is baked right into the gamma curves, the auto-exposure settings, the focus algorithms, and even the gain of the sensor.

You could argue given the average skintone among human photography subjects, this tuning makes sense, but it does leave some people out. I don't know if properly lighting and exposing photographs for people with darker skin is actually a harder problem. I have not tried. But we can be pretty certain that if the field of photography had been built from the ground up assuming we'd be photographing mostly people with dark skin, the entire technology stack would do a better job at producing legible images of their features and the contrast bias in facial recognition software would be reduced significantly.

I am not sure what you're getting at. It seems like the AI in all your examples made actual, real mistakes. They were mistakes that seemed especially bad because of historical racism but they were all mistakes nonetheless, and it ought to be retrained not to do those things. I think you know why putting images of gorillas on news stories about human criminals is not correct, or why not identifying black pedestrians is a problem, so let's look at the other two.

In the case of predicting crime in minority neighborhoods, presumably this is a network that was supposed to guide police to where they should increase enforcement. But high reported crime rates in minority neighborhoods is partly due to already high enforcement. This is classic sampling bias. Using enforcement level to guide enforcement level creates a feedback loop that hampers the goal of the AI. You want to adjust for current level of enforcement, at which point the statistical bias goes away and you get a more useful system.

In the case of mistaking black women for black men, perhaps identifying gender by appearance is ill-posed and we should reconsider the utility of such an algorithm.

Your rebuttals range from weak, to greatly outdated, to non-sequitur, and your posting history speaks for itself. You're a biased and possibly state-sponsored troll and I will not be engaging. It also seems your reputation proceeds you with other posters calling you out by name before even appearing.