Nothing "rogue". That term is just a lie by misdirection in this case. "Works as designed and that means does not always work well" would be the honest thing to say.

The unsolvable problem with AI Agents is that they always screw up sometimes. It cannot be prevented. Hence you can either accept the damage done (not always an option) or not use them for anything important (defeating their purpose).

So you are saying 90% of all people, you included, are stupid?

Pleas stop pushing lies.

Indeed. These people are DUMB. Especially (2) is a clear sign of a "constant delivery scam", where the next version is always promised to fix the increasing problems. Obviously, it never does. And also obviously, smart people notice after a while and move away.

Interesting. So a serial offender. Just like the person that gave him that pardon.

There are (multiple) existing systems that do all of those things. In fact, if you want to land in certain weather conditions you are required to use an automatic landing system. It was also observed (in a 1959 paper) that the autolanding systems of even that time were more precise than pilots.

This guy is wrong, the technical problems of automated flight are much easier than cars and have mostly been solved. The big remaining problems are non-technical ones.

From the use of the historic concept of a "Cheque" (I assume auto-correct made that into "check") I assume you are talking about US banks? For some European banks I know, they assume a large faction of phones are compromised. They do assume the compromise is more limited on a phone and they scan for malware (as much as a single app is allowed to), but basically they just use the phone as 2nd factor where an attacker would have to compromise both devices to really gain anything. But they carefully monitor the situation and larger banks have pretty good real-time fraud detection these days.

Yes, I was wondering the samer things. Quality "journalism" brought to you by a "technology journalist".

That sounds to me like a childish belief. What mobile devices get you is a second computer that is not easily associated with your first one for an attacker (if you are careful). But they are just computers. Not special in any way except for the from-factor.

I guess too many people still see tech as "magic".

The fascinating thing is that there are still actually people that think they have nothing to hide.

You know it has been firmly established when the Government does not even drive surveillance data collection anymore, but simply is just another customer of the surveillance industry. Have fun with the upcoming fascism.

In a safety-critical real-time system? Yes. Most software is good with 99.999% availability or even lower. If your equipment gets wrecked on interruptions and/or hundreds of pople die, that does not cut it anymore. Most software makers never venture into that area, but industrial control systems are a bit of a different beast than "apps".

He probably has decided that last time was the test run and that this time his exceptional skills and insights will finally make him rich. The guy is probably lying even more to himself than to his investors.

On the level of individual responsibility, yes. But on the level of wanting to keep society going, no. For the second scenario, you need to restrict fraud and scams, because society needs a relatively high level of trust to not fracture and that means that you usually need to be able to rely on the claims made by people about their products.

Of course, some people think a working society is optional. Good luck with that.

The typical claim is not that "ChatGPT is not useful". That is just the dishonest propaganda version. The usual claim is that ChatGPT is not useful or outright harmful for a lot of the application scenarios claimed and only has limited use where it actually works, which is mostly improved search and text summarization. And even there it comes with the risk of hallucinations.

That sounds a bit different from what you just claimed, doesn't it? Of course, you being an AI fan may indicate you are not smart enough to see that difference. I recommend you take the statement I just made and have some LLM type Artificial Idiot explain to you what the difference is to "ChatGPT is not useful". Because while LLM-type AI is utterly dumb, it usually can perform on this very low level, even if you cannot.