I don't agree, but a different method might have been better.

The main problem with the method used was a total lack of security. The obvious strategy would be to:

1. Force a secure connection where Notepad++ creates a tunnel using a public/private key pair, the public key being in Notepad++. This ensures that you're connecting to who you think you're connecting to. The download machine should not be directly on the Internet, nor should it be the webserver, it should be reached via a DMZed proxy where the proxy exposes just that one port and the downlod machine likewise exposes that one port to the DMZed machine.
2. You download the digitally signed installer file via the tunnel.
3. You validate the digital signature on the installer file.
4. If you're paranoid, you pull the SHA3 for the file from an independent path (https from the webserver?) and compare that as well. The webserver should also not be directly on the Internet, it too should be reached solely by a DMZed proxy. The webserver should not be able to talk to the download server and vice versa.
5. If, and only if, the signature and the hash both agree, do you run the installer.
6. You validate the digital signature on the installed binary - if there's a mismatch, you uninstall immediately.

This is not foolproof because keys can be compromised and the best security in the world can be broken. But this process makes breaking an entering a bit more of a challenge.

Emacs is an OS with built-in editor.

Vi is psychological warfare, but I'm insane enough to actually like it.

Gondor lit the beacons before it was under siege, because to do so after is far, far too late.

For the IT industry to start speculating AFTER it has lost a third of its workforce is to start debating whether to light the beacons only after a third of the city is taken.

This is a crisis that has been expected for a very long time. Long enough for you to have experience in fighting the bean counters. Sorry, but this is a mess of your own making. In more ways than one.

1. AI is good at a few basic tasks, but it is not good at being innovative or fresh. Nor is it ever going to be capable of being so, because you can't have the future in the training set. So regurgitating a few simple themes repeatedly was never going to be in the interests of humans, only in the interests of accountants (most of whom seem to have used the daleks and cybermen as a training manual on conduct) and short-term profits. Accountants don't care if a company goes belly-up, they work many accounts, so short-term profits (even if it causes medium-term collapse) are all that matter.

2. AI cannot write decent code. How could it - it was trained on Stack Overflow and abandoned github projects. But this only matters if the humans bother themselves to write reliable code. You can replace one bug-ridden pile of carp with another without users caring too much.

3. AI cannot write tightly-optimised code. But, then, I doubt most humans ever bothered to learn that skill, when they could simply instruct the user to install more RAM and a beefier CPU.

Until now, I had never realised that, with something as sophisticated as the human brain, it was possible to achieve an IQ that was not only on the imaginary number line but also a negative value.

It is depressing how this is actually the single-most informative post anyone could possibly write on the subject.

Using them for linear separation of states in order to classify results is precisely how they should be used.

More like base i. It is wholly imaginary.

But good ideas are common. The question is whether it'll get implemented and used by enough nations to make a difference.

If the EU adopts this, which is possible as there's a pressure to move away from the US, then there's a real possibility that it'll make a difference. But one nation isn't nearly enough to push things over the tipping point. Especially with the UK and US governments determined to destroy privacy and internet security.

Using tools and making tools are very different things. The only species that actually manufacture tools are crows, humans, and chimpanzees.

At any given point in time, there seem to have been 4-6 different branches of humanity, with only one of these surviving to the next round (possibly occasionally more). There have been something like 5 different rounds in the past 2.2 million years. This suggests that between 20 and 30 different lineages of hominids have gone extinct. If some of those other lineages did successfully branch, then this figure goes up accordingly.

Homo Sapiens aren't a particularly competent branch. Most of the early innovation (art, music, symbolic thought, ritual) was done by Neanderthals and only later adopted by H. Sapiens, and Neanderthals lasted around 450,000 years - H. Sapiens might well drive themselves extinct after a paltry 300,000, showing vastly inferior adaptability.

You're absolutely right that survival has been dumb luck by humans so far, and that acting intelligently and globally will be a requirement. Humans, I'm sad to say, just don't show that kind of competence.

Yet you can't provide these transcripts, nor evidence that they were genuine or, indeed, existed at all. Interesting.

A forged image is VERY VERY different from a difference in interpretation. If you can't see the difference, then I'll assume that you have no actual objection to falsified images and that you will offer no criticism of ANY such image posted by ANY side for ANY reason or purpose. Only, we both know that that won't happen. You'll criticise those you fantasise as enemies and hold triumphant those you see as allies, no matter what is done, nor to whom. You have no interest in truth, only tribes.

Frankly, you aren't a geek or a nerd, you're just an idiot.

If you want to find anything that is unverifiable or unfalsifiable about the work of Dr John H Conway's theorem, you are welcome to do so and publish a rebuttal paper. If you don't, then I will accept that as evidence that you didn't bother to check the theorem or did and found that your claim doesn't hold water.

QM works just fine with Relativity. There are two areas which give any problems, gravity and time. However, Quantum Field Theory requires neither to exist, and therefore there is no particular grounds to think that this is remotely of concern.

I use a definition I've extrapolated from the Turing Test:

In maths, if f(x)=g(x) for all x, then f=g

However, for humans, f(x) isn't a razor-thin line, it's a band that follows a normal distribution, so we have to modify this a little. AIs also don't produce an absolute result but a band that, again, follows a normal distribution. (Nobody demands identical neuron firings between humans for the same stimulus, or even by the same human for the same stimulus.)

If the band for f(x) predominantly lies within 1 std dev of the mean value of g(x) at that point, then, functionally, f(x)=g(x)

Ok, that's closer. But we haven't defined x, f, or g.

Let's extend this out. Let x over an interval be the steps during any non-random conversation. f and g are how the AI and the human mind progress during that conversation. Let x, overall, cover ALL the conversations, regardless of who with, held over some non-trivial period of time (say a year). This period has to be long, because what you want to measure is not how much the person remembers but how much the person's brain (and therefore responses) have altered as a result of each conversation, because that is a key aspect of intelligence. Memory is not nearly as important as the change in the model itself, and you need sufficient time to model those changes.

Why measure a continuity? Because intelligence is about rationally proceeding through a chain of thought, not instantaneously having a thought. An AI can certainly be capable of that, and LLMs do indeed proceed, but do they proceed in a manner that we can map within the known bounds of intelligence? That is the important question. Individual thoughts and decisions aren't important, sequence is.

There is a second aspect, which is orthogonal to this, which was raised by Dr John H Conway and, independently, by Professor Roger Penrose, which is of free will.

Conway proved, mathematically, that you cannot achieve free will through anything that is deterministic OR random, that free will has to be a fundamental property of a particle or of physics itself if humans are to possess it. He did not show they did possess it, he only showed that if free will exists, it has to be found in something within physics itself.

Penrose's argument is actually not that dissimilar in that he, too, argues it has to be a fundamental property of physics. He argues that, for free will to work, you need retrocausality - the collapsed state of any quantum system MUST precede the collapse itself in some manner. That's not actually too difficult in QM, as time is not fundamental but an emergent property and therefore the chronological order is merely what we observe and has no relationship whatsoever to the events themselves.

So far, Penrose's argument is unproven (unlike Conway's), but we have to consider the possibility that he is correct. If he is correct, then LLMs won't do the job, you'll need a quantum computer that therefore possesses the precise property needed.

Depends. If "do" is interpreted as per the Carry On films, then possibly it does.

If "do" is interpreted to mean "can actually be deemed even vaguely like", then you're absolutely right.

The big problem with fusion is the investment in it.

The International Monetary Fund estimates it costs $11 million every minute in subsidies to keep fossil fuels working.

The entire investment over the past 60 YEARS for fusion comes to around 3 days of fossil fuel investment.

We don't have to match dollar for dollar, although that would be very nice and I'm sure it would improve things a lot. Spending the equivalent of a fortnight's fossil fuel investment in fusion research (more than quadruple the total spent to date) each year for the next 5 years would almost certainly yield working fusion reactors that were extremely cheap to run.

Once you're over the crest, then you can not only cut out most of that fusion spending, but you can cut out most of that fossil fuel spending as well.