Yet. Continuous progress in robotics together with the combination of robots with AI can create machines good enough to replace a lot of manual jobs too. This may happen sooner than we think too: see how quickly AI has evolved from a subject for derision, with its drawings of six fingered people, to a real threat to knowledge workers. Fairly soon I expect truckers, taxi or Uber drivers and others to start feeling the pressure from autonomous cars.

In the end, the driving force is the desire of companies to replace expensive and unreliable human workers with machines that can be owned outright. As long as somebody gets paid to do something, they'll be a cost to the company; add in various worker rights - ragged and flimsy as they they have become, especially in America - and companies are strongly motivated to get rid of the pesky humans. Machines don't even need to be better than humans for some tasks; if they're good enough and don't cost too much more, they'll be used.

The replacement juggernaut can be faster in some areas than in others but the direction is the same everywhere. To begin with, job seekers will have to accept lower wages and worse conditions so they can compete with the advancing machines. Still, sooner or later it'll be too expensive to pay a human for virtually any job.

Obviously, given the current social system, this is unsustainable. We as a society need to separate living from employment - because there'll be no employment available for the majority of people.

Just move your kid to a local school - problem solved!

What, you say there are no good schools, because nobody pays taxes? And the only schools are private and too expensive? Geeze, I wonder how this happened!

Well, it's weird that your first thought was porn, but this is really one of the worst examples you could have chosen. First, all you need to make smut about some person is a name and a few photos, or maybe only one - and even the name is optional. Your putative malicious porn maker doesn't need any of the data the lawyer in the article targets. Your argument doesn't have any legs.

Furthermore, are you so naive as to think that AI is the only way to make porn? Do you think there was no "look-alike" porn before AI? Drawings, paintings, smutty books about various living or dead persons, weird fan fiction, more recently porn movies with models made up to look like somebody else are a dime a dozen, and have been so for centuries or more. Nobody needed AI to make them.

Also, you have nightmares about porn? Geeze!

This whole concept of "privacy for the dead" seems iffy to me, and I'm also not sure why the author singles "AI" as the particular scenario that raises concerns. People have been writing histories, newspaper articles, obituaries about dead people since forever. They wrote biographies, painted pictures or carved statue of dead people, they made movies/biopics about them, wrote "tell-all" or other unauthorized biographies without needing the estate's approval and sometimes facing explicit opposition from the estate. Is the author's intention to stop all this, or give the estate an absolute right to veto all and any of those?

Also, what exact data should be included? Should the estates have the right to delete or block access to public info about the deceased, like property records, police files, jail sentences? What about posts to social media, which were intentionally published to all and sundry by the deceased ? If somebody has read the deceased's posts, is he supposed to forget them, or never use them in any way without estate approval? Also, some social media companies explicitly state in their TOS that they reserve the rights to use the content published by the poster in any ways they want. Should the TOS be invalidated retroactively? What about if the respective company is from some other country?

One area where I find AI useful is in automated transcription and summarizing of Teams meetings.

Unfortunately, the trend in my company is to move various discussions from e-mail to Teams threads, or - even worse, IMO - to various meetings over Teams. This is principally driven by younger generation folks, who seem to think e-mail is too old-fashioned, despite Teams having the worst UI and usage model I have ever seen. This particular rant aside, AI integrated with Teams has done a lot to alleviate the pain points. AI gives me a text summary of the meeting that I can scan quickly for main points; I can then ask for clarifications, or ask the AI to give me a shortcut to the time in the meeting when some issue was discussed. AI can also build lists of work items decided in the meeting, lists of areas of disagreement with summaries of the arguments on both sides, and I get all that without having to listen to folks rambling about - more or less coherently - for minutes and wasting everybody's time.

Well, to be fair, EV opponents argue in a very similar way: "this is a problem for me so it will be a problem for you".

Well, I'm not an astronomer, so my opinions aren't in any ways authoritative; however, from what I learned, the lifetime of stars is - interestingly enough - inversely proportional to their size. Very large stars burn through their material in a mere few million years before exploding; the smallest red dwarf are misers and can live for trillions of years.

Given the current estimated age of the universe, there was plenty of time for generations of massive stars to be born, burn brightly for a few million or hundreds of millions years then explode, enriching the interstellar medium with metallic atoms ("metals" is what astronomers call any element heavier than helium, so this includes oxygen, carbon and all the rest).

This is what Sagan famously said:

Our Sun is a second- or third-generation star. All of the rocky and metallic material we stand on, the iron in our blood, the calcium in our teeth, the carbon in our genes were produced billions of years ago in the interior of a red giant star. We are made of star-stuff.

Well, that's not quite the case. As stars age, a portion of their stellar material gets dispersed in planetary nebulae. If a star becomes a supernova, he huge explosion also disperses a lot of stellar material. Even if a star collapses to a black hole, some stellar material still gets ejected via relativistic jets.

This material, which has already been part of a star, can coalesce again, creating new stars. Supernova explosions create shock waves in the interstellar gas, creating zones of high concentration, who become new star nurseries.

The first stars after the big bang were composed mainly of hydrogen and are called population III stars. During their lifetime, they created heavier elements (in astronomy-speak, "metals"). When population III stars died, they enriched the interstellar medium with those heavier elements, and the second generation of stars (population II stars) started their lives with higher metallicity. When population II stars died, the process repeated, and another generation (population I) came to live, reusing the stellar material of predecessor generations.

The Sun, for example, is a relatively recent population I star, and has comparably high metallicity.

This looks like mysticism - or, to be charitable, maybe like a reference to the quantum consciousness theory. I'm not convinced this theory is true, but I'm not knowledgeable enough to judge, nor arrogant enough to state my opinions as facts.

There are many quantum random number generators that use quantum effects to extract non-deterministic streams of random numbers. Some types are a dime a dozen (like the ones based on electronic quantum noise in reverse polarized diodes, for example). It's trivial to integrate non-deterministic quantum random number generators in an AI, so your argument doesn't hold.

Sure, but for computers, I'm not sure the separation between the two can remain as strict as it is for humans. With access to networking, the AI may run the specialized chess program as a sub-agent and access its skills - in fact, the AI can access a whole variety of expert systems, and integrate with those. In a way the sub-agents become part of the AI itself, while the top level ChatGPT or whatever component becomes the "conscious self" of the wider distributed intelligence.

I'm not a psychologist, but IIRC there are well-developed mind theories that suggest that something similar is happening in humans. For instance, Daniel Dennett's Multiple Drafts Model, or Minsky's society of minds theory propose that the human mind is built of simpler components (named agents by Minsky), who are themselves mindless.

Ook!

It shouldn't take 4 billion years; on the contrary, I believe this will be very fast. This is a problem, because I don't think society will have enough time to adapt, leading to turmoil and potentially big trouble.

The reason why I think the change will be very fast is the fundamental difference between biological evolution and technological evolution. Biological evolution uses a Darwinian model, whereas technological evolution uses a Lamarckian one. In Lamarckian evolution, the offspring inherits the physical characteristics the parent acquires during its lifetime. For example, say some giraffe ancestor starts eating leaves higher up on trees; the effort will strengthen its neck and maybe get it a little longer. Lamarck's assertion is that the offspring will be born with a slightly longer neck than the parent's.

Biological evolution happens via Mendelian genetics - acquired characteristics don't get inherited (barring some minor effects of epigenesis). This makes it SLOW. However, improvements acquired by a generation of some technology do get passed on to the next generation. The speed of technological evolution is explosive, compared to biological evolution. We have seen it already, as a single human generation saw technology evolve from horses and carts to landing on the Moon.

We see the same break-neck speed of change in AI. Just a couple of years ago we were all laughing about clumsy six-fingered figures; now, we're talking about making full-length feature movies with AI. I also think that the technology still has a long way to plateau, and I don't know what will happen along the way.

As Sir Terry Pratchett said, the world could especially do with a few more orangutans.

I think I read a document about that - they use emergency teams of children to stop leaks by sticking their fingers in the dam's cracks.

That's ridiculous. Do you somehow live under the impression that Tesla is the only electric car maker in the world? Tesla sales are declining sharply in Europe and China, while European companies like Volkswagen and BMW or Chinese brands like BYD are taking over. Tesla role as a driver of innovation has also pretty much disappeared, and they have trouble coming up with new models as they get sidetracked by technological debacles (like the Cybertruck) or by Musk's political idiocies. If Tesla closed doors tomorrow, electric cars would continue to thrive and grow - maybe even better than they do today, since the corrupt Republican administration (I know, oxymoron) wouldn't have more incentive to cripple the competition in order to favor Musk's "Tesler".

Also, blaming the left's hatred of Musk for the putative failure of electric vehicles is doubly ridiculous. Do you somehow feel conservatives are putting in sterling efforts to encourage electric cars, but are hindered by the hater lefties? Are you even listening to yourself?