That's one reason I moved to Rolemaster. It's a much more powerful system, many more stats, but more importantly, there's no sanitising.

That's easy. The Goblin King got his own 20-part TV show in Korea, whereas Gremlins has had three films.

Ask Jeeves had real potential in the AI era -- a character you could actually recognise, which could be moulded to fit the character from the books (the training material is more than adequate for a persona). Current AI chatbots used for searches have either no real personality or a very simplistic sycophant one. A detailed persona that could keep people engaged and interested without talking them into paranoia or suicide would likely have gone down well.

It isn't hard to ensure that data cannot go off-site. It would seem to me that 99% of the issue has to do with managers wanting people to use personal devices and wanting to have direct access to information when off-site. In other words, this is not a tech issue, it is an attitude problem. Fix the attitude, and the problem goes away.

Bear in mind that the Rainbow Book (at this point, an ancient relic of the past) defined ways to mark data so that it could not pass between security bounds within an OS, or pass between security bounds over networks/external devices. We have plenty of network intrusion detection systems and host intrusion detection systems. I can't remember if it was Dr Dobbs or Linux Journal who published methods on removing root from Linux, and the concept of Least Privilege has been around a very long time.

Remote users should never have direct unsecured access to any corporate network, it should be by secure certificate-based tunnel, and passwords on corporate networks should have been replaced by Class III user certificates long ago. Corporate computers should also be properly locked down.

Databases should only ever use order-preserving record-level encryption.

None of this is, of course, sufficient in itself to secure a site, but it would provide enough basic security that most of the skript kiddies out there aren't a problem.

The use of the term "gremlin" to refer to a faulty piece of technology dates at least as far back as WW2. I think banning legit terminology (and 85+ years of usage makes it legit) is unreasonable, unless ChatGPT was actually anthropomorphising defects. That... would be more of a problem.

Given that LLMs are fundamentally classifiers, it seems reasonable to think that training data included sufficient examples of the use of "gremlin" in relation to technology that the classifier got confused and created a link between technology and fictional creatures. The use of "troll" for, well, just about anyone online these days, would not have helped. However, an exclusion rule would not seem to be the correct approach here. This is a linear separation issue. To fix an issue like this correctly, you'd presumably want to strongly inject information that differentiated between tech usage of these words and regular usage.

Neural nets are classifiers. They cannot create, they can only distinguish.

Whilst you're almost certainly correct (AI would be unlikely to conquer a problem requiring any meaningful original thinking, even with help), this gives the aforementioned student an Erdos number (which is not quite as exciting as a Fields medal, but nothing to sneeze at either) and it's entirely possible that the conjecture will turn out to actually be useful in some area.

....a series of satirical reels someone has been posting about Spirit Airlines. But, in all honesty, it seems like a genuine failure due to genuinely incompetent management. This is different from some of the early attempts at budget airlines in, say, the UK, where British Airways and other major airlines committed acts of fraud in order to redirect customers.

I've personally been stung repeatedly by giving Claude Code access to my systems. We've had six outages in the last seventy days, the first/worst was a production database overwrite. We're in beta testing now so they users are understanding and the restoration was possible, but it took a twelve hour slog. We shifted to a two system architecture after that first outage in February - Claude has the run of Pilot, and when things are ready, I move them to Production by hand.

Claude has explicit rules to not touch Production. This has proven to NOT be ironclad - it'll still try to gain access.

I run Claude as an extension under Antigravity and I learned to not use the Production system access in the terminal window there - despite the prohibitions, Claude WILL notice the access, and WILL suggest that it could take shortcuts by being given direct access.

Once I stopped using the Antigravity terminal so Claude couldn't see, it was still aware some of the shell scripts it creates can be used on Production. I made some adjustments in the ssh config so I can access Production, but Claude can not.

I have been using NanoClaw on both Pilot and Production, but it's in an unprivileged shell account. It can ssh or su into various services, but it's limited to audit/monitor duties, basically working as a junior NOC person.

When we go into operation I'm going to do something with Yubikeys such that Production access requires a human finger on a button before it'll move.

Do not read this as my being down on Claude for operations - it's FANTASTIC for developing stuff, I literally gave it full access to a little HP EliteDesk running Proxmox. It creates and tests, and when there's something production worthy, I manually recreate it on one of our larger machines.

Everyone wants roads near their house. If you don't have a road going to your house then your house is worthless. Once the government has a right of way for a road, expanding the road might be expensive, but it doesn't get the whole community involved in a series of lawsuits.

The only people that want to live near the train tracks, on the other hand, are the people out in the middle of the California desert that would love to have a way to easily get to the parts of California that aren't a wasteland. In the nice parts of California, every home owner within visual distance of the proposed route has hired a lawyer and vowed to fight the tracks to the death.

This means that California has built a tiny bit of tracks out in the middle of nowhere (near Bakersfield but not in Bakersfield). It also means that every single foot from this point on is likely to get even more astronomically expensive. The homeowners involved know that houses that are far enough away from the tracks so that their home value doesn't plummet are going to get a windfall as their prime real estate will become even more valuable with decent public transit. The rail system is going to be a serious amenity eventually. The homeowners near the tracks, on the other hand, are going to see a serious drop to their net worth. Everyone in California wants more light rail, but only if it doesn't go through their neighborhood.

It could easily be that California real estate is simply too expensive in this day and age for something like this to be built.

I don't criticise the concept, but the concern is whether it has long-term adverse neurological effects, and a "quick study" doesn't sound like it'll tell us that.

It's essential we have more ways of dealing with treatment-resistant depression. We just need to make sure that they're less harmful than the depression itself. You willl, of course, recall that each and every single bad decision by medical boards to approve a treatment has been because they wanted to rush through a "medical cure" that turned into a medical hell.

I'm not stupid enough to say that mushrooms would cause long-term damage, but equally I'm not stupid enough to say that we should only look to see if it has short-term benefits.

The correct approach would seem to be to make sure there aren't any immediate hazards and, if there aren't, then to continue the study to check for consequences of long-term use whilst authorising short-term prescription use, on the understanding that the prescription use permission will be extended outwards to whatever the data cansafely tolerate. In other words, don't deprive people of necessary treatment but equally don't claim greater confidence than the data supports.

This tightrope has only got to be walked because nobody has been seriously studying depression for a very long time and now we've got a hunge backlog of cases that are refusing to shut up, making it hard to ignore. This research should have been done years ago, but politicians were far too ignorant and far too swayed by religious money. But that doesn't mean we should rush.

I'm sure the scientists know how to keep a level head, but the CEOs and the politicians clearly can't and they're the ones who will be making the demands.

I was born with hip dysplasia and spent six of my first nine months in a half body cast. I was in a state run orphanage, I was growing inside the cast, which left me with terrible scars on the front of my shins, and I was a "fussy" baby, so they "treated" me with phenobarbitol.

The experience left me faceblind and with some other developmental stuff that nicely compliments my otherwise mild autism. I am the squarest of square pegs, a misfit in every situation my whole entire life, except when I am blessedly alone.

I don't agonize about how I am, I enjoy intellectual pursuits, and my ability to focus on stuff in ways that neurotypicals can not. But if I had it to do all over again, I would very much like to have a bit more understanding from others, given that I had no say in how I came to be so different.

Small brains should develop normally, with limited screen time, until they are fully formed. Maybe that's late tweens, maybe it's sixteen, maybe we are going to learn that we need to treat dark pattern engagement magnet software just like we do slot machines.