We are developing the social individualist meta-context for the future. From the very serious to the extremely frivolous... lets see what is on the mind of the Samizdata people.
Samizdata, derived from Samizdat /n. - a system of clandestine publication of banned literature in the USSR [Russ.,= self-publishing house]
|
Samizdata quote of the day – AI edition “When sweeping, idealistic dreams trickle down into sales and marketing channels, AI’s potential uses become unclear. Framing AI as a general-purpose Swiss Army knife for productivity inevitably leads to paralysis for its end users: Where do you even start with a technology that can do everything?”
– Parmy Olsen, Bloomberg ($)
Along with others, Olsen is freaked out by the skyrocketing ascent of chip-maker Nvidia’s stock price.
|
Who Are We? The Samizdata people are a bunch of sinister and heavily armed globalist illuminati who seek to infect the entire world with the values of personal liberty and several property. Amongst our many crimes is a sense of humour and the intermittent use of British spelling.
We are also a varied group made up of social individualists, classical liberals, whigs, libertarians, extropians, futurists, ‘Porcupines’, Karl Popper fetishists, recovering neo-conservatives, crazed Ayn Rand worshipers, over-caffeinated Virginia Postrel devotees, witty Frédéric Bastiat wannabes, cypherpunks, minarchists, kritarchists and wild-eyed anarcho-capitalists from Britain, North America, Australia and Europe.
|
Presently “AI” means computers programmed to dish out leftist propaganda – often blatant lies.
Whether true Artificial Intelligence, as in a free will being (a true “intelligence” – a person) who will check things for themselves and come to their own opinions, will be created – remains to be seen.
It may be done (I do not deny it) – but it clearly has not been done yet. What has been done so far (see above) is the creation of machines who trot out the same propaganda (often blatant lies) that the rest of the establishment trots out – and the machines do this because they are programmed to do so.
One problem is that many people involved in “Artificial Intelligence” do not believe in intelligence (in personhood) they do not even believe that they themselves are intelligences – persons (i.e. are free will beings) because they do not believe that personhood (moral agency – free will) even exists. They are, sadly, utterly twisted (for example the pet “philosopher” of the World Economic Forum) and their creations reflect that.
As for the stock market bubble – well all the Credit Money being created (being created from NOTHING – nothing at all) has to go somewhere.
In real terms the stock market (and the system generally) will have collapsed by the end of 2025.
The Credit Bubble economy is not sustainable – it will go.
The present international economy is based on foundations that are both insane and evil (rotten to the core) – it can not last, so it will not last.
Being old and cynical I observe that this year’s solution to life, the universe and everything generally turns out to have a somewhat narrower application in practice. For example no matter what nonsense politicians come out with electric vehicles cannot replace ICE vehicles overall but do suit some use cases.
With a typewriter I can write *anything*! When you can write anything, where do you start?!?
With a pencil….
With the human vocal cords….
Just another bitcoin, Internet of Things, and-then-a-miracle-occurs, overhyped piece of nonsense. I apply the standard llamas credibility test – if a person who knows something about this can’t describe what it is and how it works in a single sentence, without hesitation, repetition or deviation, then chances are, it’s a crock of speculative s**t (at best), or a scam (at worst). Double if it’s described as ‘the solution for everything’, having ‘no limits and no boundaries’, being ‘the road of the future’, or (especially) how we need to get in now to avoid missing out.
We’ve all seen many, many of these super-hyped promises of endless sunshine and puppies come and go. This is just one more.
llater,
llamas
But how close are we to the “trough of disillusionment”? [see “AI Hype Cycle”]
Still many years away I suspect.
The AI of today are the learning systems of old but with much better hardware. They aren’t smart but they can be powerful. People think that some jobs will be replaced with AI but that’s not really true, in fact most people’s jobs are pointless and unproductive – just moving word salad from one place to another. Only about 5-10% of people in most organizations do any productive work. In time the AI will analyze a company, identify the productive people, connect those people together in an efficient way and then the remaining 90%+ will get the chop. So most people won’t be replaced by an AI, they’ll be fired by one. That’s what learning systems are really good at.
When will the crash come… Soonish. Then a climb-back like the one that followed the dot.com boom/bust. Now, this is where it gets interesting… Who survived that bust? Paypal & Amazon spring to mind. What do they have in common? Trad products. Amazon sells physical stuff and does TV. Paypal simply facilitates purchase and does it very nicely. Some AI will survive and prosper. I wish I knew which but I’ll suggest they’ll be doing stuff that can be monetarised successfully.
OK. If you comment here you’ll know ReCaptcha. Why are you always being asked about vehicles and street furniture? They sell that to outfits trying to do self-driving cars. Is that the killer AI app? Yes, maybe. But there is a problem and it ain’t tech but law. You get a self-driving car crashing then the question of liability is a potential nightmare. Quite frankly I dunno what the killer app for AI is going to be but it isn’t that in the foreseeable.
So, what is my experience of AI? Mainly NightCafe, other art apps (it’s great at removing backgrounds but that is a real small niche – handy for me and other web-designers but it ain’t rocking the World). NightCafe is a curious beast. Officially you can create “artistic nudes”. Realistically it is more censorious than Mary Whitehouse. Give you an example. It hates the word “frontal” regardless of context (“full-frontal nudity” – I assume is assumed). NightCafe is clearly paranoid about the possibility of getting into legal troubles if it is seen as enabling deep-fake kiddie porn or similar. Just like self-driving cars this is about the legals.
Now, enter the disruptor… AI tends to work on a server… somewhere. Just recently the likes of Dell, HP, Lenovo have revealed laptops with NPUs which means the AI is actually on the users machine. I have no idea what that means legally but it will be interesting.
So, the crash will come from the legals as much as anything. The resurgence will happen as that is sorted.
Anyone interested can check out my recreational art on NightCafe here.
llamas
I apply the standard llamas credibility test – if a person who knows something about this can’t describe what it is and how it works in a single sentence, without hesitation, repetition or deviation, then chances are, it’s a crock of speculative s**t (at best), or a scam (at worst).
I don’t agree at all. It is the nature of new technology that it is unknown and hard to understand, even for experts. Can you explain in a sentence how your phone works? I know a huge amount about how your phone works, but couldn’t explain it all in a twenty volume book. “Your phone lets you speak to people at a distance, keeping you in touch no matter where the other person is.” Is that a good description? My kids almost never talk on the phone, but use it constantly in other ways. Heck forget my kids, I almost never talk on the phone, but get the jitters if it is more than ten feet from my body.
When cell phones were first released nobody could possibly anticipate how transformative they would be. Experts were primarily concerned about how to use bandwidth as efficiently as possible, and to minimize power draw. I remember when they first started putting cameras in cell phones and I was like WTF? I mean in retrospect is was brilliant, but it was far from obvious that that was a good idea — especially with how crappy the early ones were. But that one sensor has probably done more to transform the world than anything else in your pocket.
When micro computers were first released and people got them in their homes a lot of people said “what is that for”, and they were justified in asking that of the early ones — you couldn’t even use them to write a letter. Pretty much the only thing you could do with early computers was play games. Then someone plugged in a modem…
So to be unsure about what a technology can do and what its significance is or how it works does not at all mean it is hype, it just means it is new and unclear and only time will tell.
This is further complicated by the fact that Ai has been lumped in the public’s mind to mean “ChatGPT”. For sure these LLMs and extremely important but AI is a lot of other things built on top of similar technologies, like voice recognition and production, computer vision, generation of images, analysis of images and so forth. It is simply a fact that AI is transforming medicine. AI systems are better at reading X-Rays and many other types of scans than radiologists, and they will continue to get better. But that is clearly just the start of a locomotive crashing through both medicine and every other knowledge based skillset.
ChatGPT is transformative in the same way graphical user interfaces were with the Mac and later Windows in the 1980s, and for the same reason. The make the underlying technologies more accessible, but there is a lot more to AI than LLMs.
@NickM your artwork is really fabulous. I was looking at the images of the pretty princess type girls and they are amazing, but they are just dipping into the uncanny valley a little, and I’m not quite sure why. I think perhaps the reason is that they are too perfect, and real people aren’t. Maybe a few very small flaws would make this indistinguishable from a photo. The little dwarf guy is really amazing.
The one with the young girl surrounded by crayons — is that a photo? I mean it sure looks like it.
And for people who think AI is a damp squib… you might want to look at this. When you really can’t tell what is a photo, and later what is a video, how can you really know much of anything of the world outside your personal experience?
Thanks Fraser! It really isn’t a photo or based on one (you can upload photos and manipulate them). My pics with crayons came from an in-joke with one of my fans on NightCafe. It doesn’t even use photo-style prompts. In fact quite a bit of my stuff is probably very difficult to figure unless you know the backstory. The raccoons are an example. It was an AI art duel. This guy did a smoking raccoon so I hit back. It rapidly went (literally) nuclear. Hence the raccoon in the deckchair with an umbrella drink and a nuclear test in the background. That was retaliated against so I did a Darth Raccoon on the Death Star and eventually Zeus Raccoon hurling lightning. The thing is it is enormous fun and fits in nicely with those spare couple of minutes…
Too perfect… I think you have a point. But art (of all forms) tends to seek beauty and a huge part of how things like NightCafe operate is their computer models being refined by competitions and upvotes. I haven’t spent a penny (there is a subscription option) because I do things like vote and sometimes do OK in competitions to get credits.
I think the creation of the beautiful via AI is because that is what millions of users desire. AI art, from my experience of NightCafe, is very reflective of the users ideals. MS Co-pilot… I dunno. I have tried it. And it is much more skewed towards MS views. I’m not 100% sure because I haven’t used it much because the UI is a confused mess.
Anyway, thanks. And if you wanna try NC out please mention @NickM. Gets me credits. I think it might get you some as well.
Paul
I think that this analysis over-complicates the situation. Most simply do not analyse what they think or contemplate the sources of their motivations. Does you average Leftist reject free will as a concept? I doubt it, they do not see people as ‘flesh robots’ but on the contrary the very free will that they see, a will not bent to their ways, is what they hate. They hate free will precisely because it means resistance to them or those they approve of.
The talk of people as ‘flesh robots’ etc. is just cover for the murderous instincts of those who know they are actively doing evil.
Mr Ed.
The popular philosophers, for example the pets of the World Economic Forum, say they do not believe in free will (moral agency) – whether they are lying to cover-up their own wickedness, I do not know.
There is also, indeed, the opposite mistake – not denying the existence of human personhood (free will – moral agency), but rather the mistake of believing in its omnipotence.
“I exist” true, “therefore the universe must be as I want it to be” NOT true.
It is not just the human person (the “I” – free will, moral responsibility) that exists – the objective physical universe ALSO exists, persons (free will agents) can not change everything just be willing it to be so.
In the end all we control our own actions (as you know the criminal law depends on people having the capacity to do other than they do – otherwise, if human actions are predetermined rather than chosen, the concept of “crime” is meaningless) – we can NOT, for example, turn the sun into a flower just by willing it so.
In places such as California government (and the vast, Credit Money fuelled, Corporations as well) is controlled by people who think that the laws of economics, even the laws of biology and physics are subjective – people who think, and act, as if objective reality does not exist – people who assume that only their will (what they want) exists.
This is why California and other such places are going to collapse.
People such as Governor Gavin Newsom or Bob Iger (Disney Corporations – which owns ABC television) do not, as far as I know, deny the existence of moral responsibility (free will) – rather they seem to be believe that their will is omnipotent, that the laws of economics, indeed the laws of biology and physics, do not exist if they do not want them to exist – and they govern (both government and corporations) accordingly.
In short they do not believe themselves to be flesh robots, but they do not believe themselves to be human beings either – they believe themselves to be Gods.
And they are not Gods – as will soon be obvious.
“The Gods of the Copybook Headings” (what Kipling called the basic laws of reason) are going to bite back – an this entire sick system is going to collapse by the end of 2025.
This is pretty good:
https://www.youtube.com/watch?v=nGIpdiQrFDU
If that link doesn’t work, search on Youtube for “Linus Tech Tips AI”.
For those here who actually know something about AI (not me!), what do you think of this video?
GregWA: I know something about AI, and even chatted with Geoff Hinton a few times 🙂 but am now behind the times.
It seems to me that Linus (or whoever it is who is talking) got the main idea, that “generative AI” can interpolate between previous data to solve problems by analogy, but cannot extrapolate to solve completely novel problems.
He says it explicitly, but it is also what he means when he says that “generative AI” and similar systems are ML (machine learning) rather than true AI.
I feel that the video was a bit too long, but i hope (in all modesty, which is unusual for me) that the above paragraph will help the viewer focus on (what i see as) the main point.
What most people are thinking of when they say “Artificial Intelligence” isn’t what’s currently on offer. What is on offer is another toolset of questionable utility, which has yet to evolve into something truly useful.
Right now, you can’t trust it. The various answers proffered on the search sites and other public interfaces are fully of error and outright fabrication, because the sources of information used to train the AI/LLM are poisoned. I’m not sure how well an honestly trained one would be, but I suspect that there are a lot of things that you’re going to have to weed out, even with the best.
Whole thing is going to replicate every one of the boom/bust cycles we keep going through, with each of these fads. I don’t see Vernor Vinge’s “Singularity” idea having much credence, these days. I could be wrong, and I might even welcome our new machine overlords, so long as they do something with these infinitely stupid satraps we’ve got running things around here today…
NickM:
(This question will demonstrate my absolute ignorance of AI, I suspect.)
When you go to generate a pic, what’s the interface? Are you simply giving a text request (“generate a pic of a raccoon smoking a spliff and watching a zeppelin burn”)?
It really is the case that the international establishment elite (both government, corporate and academic) do not believe that the basic laws of economics, and of things such as biology, are objective – they reject objective reality when it conflicts with what they want (their desires).
In short the international establishment elite is irrational (they have rejected reason) – and the machines they create reflect their own irrationality.
If asked to show C02 their machines will show pictures of smoke coming out of factories and so on – even though C02 is an invisible gas. Their machines will also claim that opponents have criminal convictions when they do not (even making up fake news stories), and-so-on.
Do you think California is well governed? Do you think it is governed in accord with the laws of reason?
If your answer is “no” then beware of “AI” machines – as they are programmed by “educated” people with the same attitudes as those people who govern California.
They reject objective truth, they reject the laws or reason, they believe only in POWER – their own unlimited power to control everything according to their whims.
That is, and has always been, the core of the international Progressive movement – and is at the heart of its rejection of what Kipling called “The Gods of the Copybook Headings” – the basic laws of reason and of objective reality.
AI is a marketing term.
Just like “cloud”, you start with a client server model, then move the server to someone else’s property, and suddenly it’s “cloud”, it’s exactly the same server with a little label saying “cloud” on it.
Bizarrely, when the security implications of having your IT kit in someone else’s basement becomes blatantly apparent, they move the server back to your own premises and call it “private cloud”.
As I work longer in IT I meet more and more people who don’t actually understand IT, exactly the ones gullible enough to adopt sales slogans as fact.
Yeah, it’s the machine learning and pattern recognition of old, with better hardware to make it real.