I've read studies that claim that homework in general does not actually help students achieve more. Those who need extra practice and help the most are the least likely to have said help at home, making homework actually counterproductive.
So in this case, I propose the following:
1. Take a step back in networking students. Stop issuing computers, tablets, and such, at least internet connected ones.
2. Stop assigning homework.
3. I recently found out that daycare costs as much as school. In some cases more, some cases less. Most people work all year, so eliminate summer break and just have more shorter breaks. This allows the work previously done as homework to instead be done in the schools, and prevents the loss of progress our long summer break currently has - it takes weeks to just catch back up to where students were before it started.
While it won't help everybody, it might reverse the decline we're seeing right now.

California already passed the exact same law last year, and Newsom signed it into law. Applies to Linux and all OSes.

If Mozilla really wanted to benefit the world, it would collect statistics on how many Firefox users deliberately disabled the AI features and publish them, or even aggressively market the results.

Why do you have a hard on for rsilvergun?

Would you give a human assistant the login and password to your inbox? Or would you set up a shadow inbox that mirrors your actual inbox so that you don't need to share your login and password?

In a similar vein, when testing automation code, do you just give it admin level prod credentials and then YOLO it, or do you create a test environment that shadows the data from prod, so that you have a way to validate what the automation code is doing without accidentally damaging prod?

Fundamental rules people! Least privileged access to do the work needed. Safeguards commensurate with the negative consequences of failures. In other words... basic risk management.

To give a slightly different example, would you let your self-installed, open source AI self driving interface (see comma.ai) drive you on the highway without sitting in the driver's seat with hands on the wheel, feet on the pedals, just because it managed to complete a test course with flying colors?

The example given with regards to the openclaw agent is like sitting in the back seat of that self driving car, then desperately trying to climb into the front seat when you realize the AI driver is about to drive you off a pier into the ocean.

The original choice was a woman, too. This chick was a former MS employee before she went over to Instacart. She's 100% AI focused. I'd say "RIP XBox," but that was probably already the case.

There are three problems when dealing with legacy code.

1. Figuring out what the code does.
2. Figuring out what the code was supposed to do.
3. Figuring out what the code actually should be doing.

The three are often not the same. The code lies. The comments lie. The commit messages lie. The documentation lies. The managers lie. The users lie.

By lie, I mean, what they tell you, regardless of what they believe to be the truth, is not reality.

For example:

Someone took a stab at writing some code in a modular fashion, or someone before you refactored it. There's a function - it says getXYZ, and it returns a value. Great! Then you dig deeper and discover that getXYZ sets several flags which are then used by the calls that come after getXYZ in the block you are looking at. You discover this only after shit starts breaking because you reordered several function calls during refactoring, none of which had the singular result of getXYZ as a dependency.

An even more straightforward example of that would be discovering a bunch of shit broke when you looked at and found that nobody used the result of getXYZ, and refactored out what looked like dead code. Again, because getXYZ, despite the pattern, actually had side effects.

At this point, now you have a problem. Is getXYZ actually supped to return a result that someone is supposed to use? Was that its original utility, and someone just jammed shit into it because it was faster than refactoring it into something else? Or was it even worse, and this was an incomplete refactor?

Nobody knows! Nobody can tell you! The commit history doesn't go back that far, and even if it did, nobody actually leaves coherent, useful commit messages!

And don't get me started on documentation and comments. Sometimes they can tell you how the system was supposed to behave at one point... but that's not how the system behaves now, and it isn't how all the users and managers believe the system is supposed to work because they've been using the current system for so long.

"Fixing" the code to follow what was supposed to be the correct design can cause all sorts of problems with downstream processes that rely on the current broken behavior. I'm going to steal Uncle Bob's example of finally fixing a typo in a dropdown menu and causing a bunch of UI macro code that looked for that typo to fail...

Often times modernization means essentially re-negotiating all the contracts and interfaces and process workflow with all the stakeholders to come up with a common understanding of what the code should be doing. That's like the best case scenario.

The worst case scenario is they say - use the old code for requirements, make it work exactly like that. Well, if the old code is shitty and illogical, and you need the new code to interface 1:1 with everything that plugged into that... well, guess what? You're going to get an architecture that is going to replicate shitty and illogical 1:1. The actual code might be great, but the process will be just as hard to understand, and probably eventually just as head scratchingly difficult to modify and maintain.

I wish our robot overlords the best of luck with this problem.

I don't know, a bus or train all to myself seems more wasteful than the regular car.

It doesn't have to be perfect to still have an effect. Though yes, I'd argue that our current system is already good enough that for like 99% of the murders remaining, serious thought on the consequences did not occur. They thought they'd never be caught or even just didn't care in the heat of the moment.

Also, getting as specific as this crime isn't as necessary. If it deters somebody from murdering their fully adult spouse or even their drug dealer, good enough.

You’ve shifted my point into something I never argued.

Deterrence doesn’t require me personally to have almost committed murder. It’s a populationlevel effect observed across criminology: when the state reliably investigates and solves serious crimes, the expected cost of committing those crimes rises, and some fraction of wouldbe offenders change their behavior. That’s true even if neither you nor I were ever in that category.

You’re also treating “people who commit murder” as psychologically identical to “people who would never consider it,” which is exactly why personal anecdotes aren’t the right tool here. The question is whether solving murders—even old ones—reduces the incentive for planned, intentional killings. The evidence suggests it does.

So if you want to engage with what I actually said, the discussion is about deterrence, institutional legitimacy, and the social value of solving serious crimes—not whether I’ve ever personally been on the verge of stabbing someone.

I think that you're incorrect, that this WILL deter others, by giving the impression that we will catch them eventually if they commit murder.

There's also the idea that the criminal justice system in general pursuing crimes even if it takes a long time for the most serious ones, helps prevent people attempting vigilante justice.

The especially concerning/scary thing this time is that what the giant corporations want is to make computing seem like "magic." Make a wish into the wishing well that is AI, and what you will receive will be what you wished for ... provided, of course, you keep paying the corporation for the privilege of having your wishes granted.

Never mind having the actual skill, talent, understanding, etc. to make your wishes come true yourself. Just pay, wish, and it will be yours ... and never mind anyone who tells you it used to be possible to get what you want to achieve without paying a giant corporation. Just keep wishing, lean how to wish big, and your wishes will come true.

This seems like the antithesis of how anyone who considers themselves an educator should think.

And the really sad part is they're not just saying this to CS students. They're saying it to writers and journalists, artists, musicians ... basically anyone whose job doesn't involve a hammer, a shovel, or a stove.

I remember cases of them digging out old bacteria samples from things like old wells, a couple centuries old, not 13k, but still resistant to a raft of modern antibiotics, more than many modern strains.
The easiest explanation is that we got most of our antibiotics by examining molds and such, and it isn't like mold and bacteria haven't been fighting for millennia already. The bacteria probably just encountered something similar enough to the modern synthetic antibiotics and had to adapt.

Ouch, I can definitely see wanting to fix the color/colour thing for consistency. Reminds me of a game on steam with ONE broken achievement. Digging into it, the developer misspelled the achievement originally - then on the LAST update, fixed the spelling in the code, but not in the hook. one character edited in binary and the achievement popped.
But I'd think that an alias would work - allow people expecting color to spell it that way, but not break already developed apps that used the old colour.

Thesis 1:

Cybersecurity companies are bloated and had a stock valuation premium created by insurance mandate (thou shalt contract with a cybersecurity company to keep your insurance premiums low) that will be going away.

Thesis 2:

People are freaking out, without basis, that #1 is true, when in fact the opposite is true - even with AI making code more secure, you will still need cybersecurity insurance, and the insurer is still going to mandate that you contract with an existing cybersecurity company in order to keep your premiums low, due to reinsurance rules. In fact, because of dumbshits using vibecoding, AND the use of automated tools to identify and chain vulnerabilities, domain specific expertise provided by a deep bench will be needed in the future.

Thesis 3:

Cybersecurity companies will be trimming headcount and employing more AI tools internally.

Thesis 4:

Instead of hiring a cybersecurity company, companies will staff their own cybersecurity departments.

Of all of these, I think #4 (companies growing their own cybersecurity departments) is the least likely. #3 is highly likely (there will be some reorganizing and continued adoption of automated tooling). And while #1 (companies will no longer be able to command a large premium) may be true in some cases, I think #2 (this is a giant overreaction, and the use of automated exploit chaining means you need more expertise in defense) is probably the most likely outcome. Building a system to ensure your code is foolproof just breeds bigger fools.