halomru

>Got a problem with that? You can deal in cash or bitcoins then.

That's an incomplete solution. Cash and Bitcoin are still subject to money laundering regulations. You just reduce the number of instances where they are applied, but you can't completely avoid them

Even computer science would suffer from pure online learning: learning to effectively work in groups, and to design and implement software together is a skill that is trained at university. Remote collaboration is entirely different and less applicable. Add to that the loss of any opportunity for lab work for embedded computing.

A lot of natural processes produce results that are normally distributed. So given a large sample, a continuous value range, no bounds (or far away bounds) and a lack of obvious skew the intuition "half the population is below average" has a good chance to be (approximately) true.

Of course IQ is a made up thing that we just defined to behave that way.

I agree it's very unlikely that sabotage is what's going on, but what you're saying is the one compelling argument in favor of sabotage. "Everyone but that one actor has very bad luck" is exactly what sabotage looks like.

Any open source project can allow you to verify binaries by making builds reproducible. The fact that most apps don't do this is indeed a security problem, but one that's far from unfixable.

With many other providers I can make deductions about likely threat scenarios. For example I know about both my ISP and about Google that it is in their best interest to serve me correct DNS entries: they have no obvious motive to do otherwise. On the other hand I can be fairly certain that my DNS queries will be fed into Google's database if I use their DNS: they have an obvious motive, have the capability and no obvious disincentive. Etc.

With some new, unknown service with unknown associations I can't reason much about the threat model.

While very useful, you can only construct a collision-free hash function if you know all possible inputs. Otherwise perfect hash functions can only give guarantees over the frequency of collisions.

In the more general case, for a hash function with n bits output, the pigeon hole principle demands that we have a collision at least every 2^n inputs.

There isn't really a single universally accepted list of continents. Some count Australia as continent, some count it as part of Oceania. And some define the continent Australia as the country Australia plus all those islands that would otherwise be Oceania.

A 100% lethal virus with high incubation period. Bunkers typically have a lot more contact with the outside than nuclear subs

Interstate spending can easily be viewed as defense spending. After all it's called the "Dwight D. Eisenhower National System of Interstate and Defense Highways" and was partly inspired by Nazi Germany's Autobahn system that was instrumental for fast troop movement.

An inconsiderate asshole is completely fair if he treats everyone equally.

That's a real problem that many attempts to quantify discrimination overlook.

Only an option in the US.

There are plenty of services that put my job in a queue and simply update the page via JavaScript to show progress and deliver the end result. The better implementations identify me via session cookie and recognize me when I check back half an hour later to get my results. That's a much nicer experience than creating another account at some random service that I might never visit again.

>Can you really fix these people? Does reminding them to do their jobs and be nice actually correct their behavior? I have serious doubts.

Everyone has a bad day occasionally. But those that are routinely bad at being nice should be encouraged to change, and failing that should be exchanged for people who will be nice and helpful. Helping customers is a skillset like any other: some are more qualified than others, and if you're bad at it there are plenty jobs for you that don't require that skillset.

That's easy enough to solve: you are only allowed to defend your SSID if you have taken reasonable steps to minimise the chance of a random name collision, and if this is really the most reasonable answer given your threat model.

For example you can't block all "Netgear WLAN" hotspots, but you can block all "Hotel Sunrise NY Guests" networks. And you can not block SSIDs similar to your private network simply because you are too lazy to specify the routers MAC in all your five devices.

After all, right now the perceived norm is to send all your 100Mb/day of data through a kafkaesk array of message queues, stream processors and map-reduce jobs.

I agree that disclosing all my bank transactions would be active surveillance and would violate privacy (even more so in countries like the US where cash is phasing out of use).

But to my knowledge that's not what actually happens in western countries. Data reported to government agencies seems to always be limited to aggregates or data that acts like an aggregate (the amount of interest earned). I have a hard time seeing a privacy violation in that, since you provide a much more detailed version of that data to the government in your tax report.

On top of that there's of course also the transactions that are reported to the government because they look like money laundering. That's more questionable, but I think it's still reasonable.

In that case Jekyll or Middleman are not the implementations you want, but static site generators still work. There are various solutions with WordPress-like admin backends that produce static pages for the frontend.

>with just two relatively low resolution cameras...

We probably can, but most of us drive with two cameras with approximately 120 megapixels of feature resolution with highly optimized non uniform pixel density and incredible dynamic range. Micromovements allow the software to boost the effective resolution by about a factor of four, to over 400 megapixels. Image stabilization works in multiple layers and executes stabilization movements within 10ms in the fastest layer.

The human eyes and vision system are certainly one of our more impressive parts, and we have a really hard time replicating their performance.

I feel like all of those questions have existing answers in the automotive industry. Modern cars include so much software that there are very few legal angles that full autonomy adds.

>Should self driving car algorithms be standardized across the industry for safety?

Just like every other safety relevant feature of my car; Baseline requirements and tests standardised, implementation open to innovation and competition.

>Because of the nature of saving lives, should self driving software be open for inspection and review in ways other software isn't?

Same rules as the software of my ABS system, or of my airbag: it's up to the manufacturer, in practise nobody cares.

>If people modify the software in the car, who is responsible?

If I modify my brakes or my steering system I'm responsible for any damages caused by this modification. Why would software be different.

>Can car "owners" legally modify code when safety and lives are of concern?

Can I legally modify my breaking systems? Yes, but it has to fulfill the same road safety regulations, and frostin depending on what I do getting all required certifications might be impractical.

I have problems with this as well. Sometimes there are obvious things that can be tested on the resulting image. For example when applying a blur on a grey image with one white pixel in the middle, I know how I expect the sum of the pixel values to change, and I know the upper and lower bounds of valid pixel values.

After the fact unit tests that verify that abs(expectedImage - resultingImage) < e are also helpful. But unless you write some modeling application that has to produce exact results, a lot of code is typically about what looks good, not about what's mathematically correct. Those parts are probably impossible to properly test automatically.

>Guess it has something to do with the change in software UI language.

In the sense that software UI follows our societies general trends. And as always in design trends, there's a feedback loop.

And plent of games and digitally shot movies add film grain.

That explains the current state, but has little prediction value. The availability of parsable data can change at any moment either because of regulation or through market forces (either external companies providing parsable data as a service, or enough companies realising that providing plentiful parsable data about themselves attracts capital).

Once enough data is available, common sense and gut feeling can be programmed as a combination of fixed rules and machine learning.

That is useful in the US, but in Europe boring but useful databases are still covered by copyright.

It's what invariably happens when you incorporate words like "new", "super" or "high definition" in the name of something. I'm still hoping that at some point our industry learns that lesson.

I would have said that SQL injection is caused by emitting unescaped user input to your SQL server

If we had any idea how they work (if they work) we could work on optimised designs that generate more thrust and are thus easier to study. But we're not yet anywhere near the stage and instead have to study a design that if it works, is so inefficient that we can hardly tell whether it does anything. Very frustrating indeed

You still give that channel your watch time, making the YouTube algorithm recommend that channels videos more

By the same line of reasoning it should be impossible to derive ethics from thought experiments.