Yep, that one, that you could still have opted to not sign, and your record would be an "anonymous" John Doe with a "don't use this" flag.

Funny thing... Once it's "anonymous", you pretty much lose your HIPAA rights, and it's even easier for a third party to get your information (because it's not your information anymore - it's just some anonymous patient), and if your case is even slightly unusual, you can probably be identified again with a high degree of confidence.

If you're concerned about your privacy, sign the form, and let people like me take care of properly protecting it. I won't say we're perfect at it, but at least we had to go through compliance training to use the "dangerous" full data set...

It's not about the investigation.

It's a thinly-veiled attempt to say "look at the valuable evidence!" and push legislation undermining all public encryption with a government-held backdoor. That, in turn, undermines the privacy-supporting modern interpretations of the first and fourth amendments, which are often distasteful to authoritarians.

For those who've never faced oppression, it's certainly appealing: The government is full of good people with America's best interests at heart, and they'll be the only ones to see the private thoughts of individuals, so they can ensure that only good and legal things are being discussed.

For those of us who've seen humanity's uglier side, the picture's a lot darker: Even good people abuse power with the best of intentions, but enabling catastrophic consequences. The only way to avoid that is to require strict adherence to a fair process.

Since the universal nature of an encryption backdoor is not conducive to that process adherence (as anyone who acquires the backdoor key becomes unfairly powerful without needing to follow the process), an encryption backdoor is as unacceptable as oppression itself.

As someone who used to be on the reading side of these arrangements, I can assure you it's all in the waivers you sign prior to receiving care.

HIPAA is really easy to work inside. Essentially, you just have to promise you'll protect the data, and you can get third-party access.

...by which time, the automated systems would have already burned excessive amounts of maneuvering fuel, and docking with the space station would be unsafe... just as happened in the test flight. The docking part of the mission (which was actually a secondary objective to testing the orbit & reentry capabilities) would still have been a failure, but instead there would have been humans in the capsule getting tumbled around.

The only failure of Apollo 13 was that a bit of wiring got damaged in a test. The only failure of the Challenger was that an O-ring was cold. The only failure of the Columbia was some insulation fell off.

The same people are designing the corporate and work environment for both projects, and establishing the standards for engineering and testing rigor. Those same people make decisions about whether to prioritize safety or schedule, and they decide how much of the corporate budget gets spent on silly things like "ethics compliance".

Yeah, that's a bit of an impact.

...And what's wrong with that?

If I have any reason to talk about an email vendor's size, it's probably because I'm talking about what they can afford to do. In Google's case, they have the market dominance and cashflow to drive new email standards, and GMail holds enough of their customer base that I can safely bet on GMail service being around for a while. That's in comparison to, say, ProtonMail, which holds a tiny market share, and could evaporate if their funding ever runs up short.

Tesla's doesn't hold the market dominance, but they definitely have the cash (and investor potential) available to support whatever project they want. If I have any reason to talk about how big Tesla is as a company, it's to discuss what the company can do. Tesla can now do things other auto manufacturers can't. That's certainly newsworthy.

Hah! Foiled by my poor typing skills and lack of proofreading... That should have been 115%, of course...

Oh, but the victim won't pay the crooks. Companies who don't want to have a PR disaster publicly say "we won't pay", then go hire a ransomware recovery firm for 11% of the ransom, and that firm turns around and pays the ransom. The company keeps their hands clean, and just says "our partnership was successfully able to restore operations".

Of course, the crooks still got paid, so they properly delete their extortion database, and everybody comes out looking good. If they were to release the database anyway, that undermines the threat for the next victim. Their goal is to get paid, not to damage companies.

Ironically enough, that's exactly what BitCoin's public ledger does. It just moves the identification problem from being in the record to needing external mapping, but that's what authoritarian regimes do best.

This is not a company problem. This is an industry problem. For decades, "IT guys" have sold security solutions as something to be bolted on to an existing system.

...and none of that will have any significant impact on a corporate threat model, but it'll cost a lot and make your CTO look like they've been really busy. Buying that much consultant time will make your company look exciting, too, but it won't protect your data.

Security is a process of ongoing improvement. It requires having an IT team that works with the rest of the business and learns what the business needs to get the job done. IT should then make changes in the order of least impact to the business. If users need to use USB drives to move data, order company-provided encrypted ones. If users need administrator access to their computers, start rolling out access restrictions so their elevated access is limited to their own machine.

As more small changes happen, more user problems will surface. That's when products are purchased - not to restrict the business, but to enable using capabilities securely. Above all else, remember it is almost never the user's fault if something goes wrong. If some piece of information technology doesn't work, it's the IT department's problem. Usually, it's because the IT guys didn't understand the business needs, or didn't provide adequate training or intuitive systems. Users will always do what's easiest to meet their goals. It's your job to make sure the easiest path is the right one.

Here, let me fix that summary for you...

See? Wasn't too hard.

The lifecycle of computing technology is no different from automobiles, firearms, horses, agriculture, or any other major technology. First it gains acceptance, then ubiquity, then advancement, before finally being assumed as a base state as the next big technology takes the public focus. These stages roughly coincide with each human generation, as children grow up with a technology to become adults unimpressed by the status quo.

Not really an article, but my own observations from spending a few months getting back into the hardware market:

Apparently the GPU vendors are similar in benchmark scores, though AMD has fewer SKUs.

Unlike the Intel/AMD competition, the prices seem to be on par for the performance, so purchase decisions are more based on features. The green guys have done quite a lot of nice things with computation (CUDA, Tensor Cores) and ray-tracing... but most applications (including games) aren't written to take advantage of those features yet.

We seem to be at something of a plateau in gaming graphics technology, with Nvidia taking the aforementioned first steps toward future progress. In the red corner, AMD's latest advancements (like in TFA) have been driven by putting multiple CPU cores on a single "core complex" (CCX), then combining several CCXs into a processor. If that technology can be applied to their RDNA GPUs, AMD might produce a single card with multiple onboard GPUs and caches driven by high-speed memory buses, essentially acting like CrossFire/SLI on a single card. They might not get as many shader cores as Nvidia's offerings, but each shader's performance could be significantly higher, or the shaders could be assigned in groups to do different renders in parallel.

There's a third player in the game that should be considered as well: Microsoft. For decades, DirectX compatibility has moved hand-in-hand with GPU technology. DirectX 12 is about 5 years old now, and though I don't know the API myself, I doubt it has much support for the upcoming GPU technologies. Also notably, the newest XBox version runs on AMD hardware, as well, so if AMD is going to try to make a shift in the GPU market, they could leverage that partnership, driving a new DirectX version. Nvidia would then answer with a new round of cards compatible with DirectX-probably-not-13.

That's all speculation, though. The reality is that Nvidia is the only GPU vendor selling specific ray-tracing or AI processing capability at the moment, and without a de-facto standard API like DirectX, they get a huge first-mover advantage to decide how those technologies are mareted. A low-cost RTX card could make ray-traced graphics a standard expectation in the marketplace, and AMD would be stuck trying to catch up for the next decade or so.

Whatever the new features are, and whenever they arrive, I'd expect the next generation of triple-A games after that will be able to start taking advantage of all that new tech, and how well those titles sell will determine what features survive in consumer-grade GPUs. Personally, I think we're right at the end of AMD & Nvidia being essentially equal for most consumer market segments, but I don't know who's going to come out victorious in the next few years. Nvidia's definitely fired the first shots, but the battle's just beginning.

The reviews don't bother me much. Unless there's a long list of negative reviews, I tend to assume good faith on the seller's part, and I don't recall that ever ending in calamity.

On the other hand, Amazon's Q&A section is often the most disheartening section of the product page:

Thank you, you insufferable muppet. You have the intelligence of a dead hamster, and appear to be most useful as an object lesson in the importance of birth control.

I'm not aware of any profit motive behind posting answers, so I just don't understand why someone would waste their time. Sure, I've seen some obvious bots posting "This product doesn't do $BASIC_FUNCTION, so I bought $UNRELATED_PRODUCT instead and it worked great!", but these aren't even spam... they're just useless answers.

Apparently the author has either never seen a CubeSat, or has never seen a briefcase.

From the full report:

It kinda strikes me as "these companies didn't lobby enough".

It would seem to me that if you're going to go to the effort to set up a driving test across thousands of miles, you'd at least check on everybody while you're at it. Maybe you don't aim for particular regions to check the unsuspected carriers, but at least do some due diligence...

That said, this is also in the report:

The recommendations do suggest changing the process to more thoroughly verify every carrier's maps.

This is expected. It's what usually happens when a group starts thinking that idealism is more important than ideals.

Large companies (and Google in its younger and smaller days) like to talk about "ethics" and "responsibility" and all those things people love... but they rarely lay out exactly how each project fits their definition of ethics.

Eventually, some controversy comes up, and the employees try to reconcile it according to their idea of ethics. They accuse the leadership of seeking profits instead of following the undefined ideals. In turn, the leadership accuses the employees of missing the corporate culture. The divide grows until eventually, the company is no longer able to function. It seems Google has now suffered one of the common first symptoms of that divide: leaks of confidential information. The employees don't trust their leadership, so they also don't respect the leadership's request to keep secret things secret. All the NDAs in the world won't help if there's a lack of respect.

Now the leadership faces a choice. They can either crack down on misbehaving employees, applying a heavy hand to any transgressions and hoping for selection pressure to keep their corporate culture where they want it, or they can try to restore the respectful culture. They seem to have chosen the latter, in this case by breaking the massive company into smaller teams and reducing the interaction with the company as a whole. That means employees have more interaction with their immediate leadership and feel more loyalty to the team. The downside to this approach is a loss of loyalty to the company as a whole, and that's the other half of this change. By ending the "Don't Be Evil" era, the leadership manages the employees' expectations. They won't be disappointed by an unethical project if they weren't expecting ethics to begin with.

Yes, it's a sad day for aspiring idealist techies. It's also unfortunately predictable as a long-term effect of having structure that's too loosely-defined, in an organization that demands adhering to a monoculture.

Yup... and that's why so many products are covered with "no user-serviceable parts inside" labels.

Liability is a huge legal issue, and gets very complicated, very quickly. The issue of liability (and ways to avoid it) has been one of the driving factors of Western civilization for the past millennium or so.

The only thing that sets Apple apart is they have a lot more money available, making them a lot more appealing to anyone looking for a quickly-settled lawsuit.