• S&T Moderators: streaM Freak

Is A.I. waiting to take over ?

Stating that a heirachy exists does not infer that appropriate logic was used to create that heirachy.

sure. it doesn't imply it either :)

but i don't believe i argued that it does.

Understading is the process of applying the appropriate logic when classifying information.

great. so if you take that definition of understanding and your definition of intelligence ("Intelligence is understanding how to classify a statement as wisdom, knowledge, information or data.") and combine them, your definition of intelligence is:

"Intelligence is the process of applying the appropriate logic when classifying a statement as wisdom, knowledge, information or data."

i'd say, by that definition, a.i. is certainly intelligent.

alasdair
 
Well, we will agree to differ.

I have always thought of AI heirachy being akin to naïve set theory which can give rise to paradoxes. I know Russell used the simplest example to explain his paradox but the problem isn't limited to that one example.

The way that the paradox was removed was by removing unrestricted comprehention and the introduction of the axiom of choice.

A far more mundane issue for me is that natural languages can produce syntactically perfect yet ambiguous sentences which is a product of natural languages themselves not being precice. I'm not pointing at AI on that. I'm only too well aquainted with examples written by people which in theory should have been checked by an editor.

I believe long ago I said one major improvement would be for AI to push back against ambiguity. So I've not just invented it just now to support my view.
 
i can agree to differ but i would love to hear why you believe that a.i. is not capable of "applying the appropriate logic when classifying a statement as wisdom, knowledge, information or data."

i'd argue that a.i. is not just capable of it, it actually excels at it.

using your terms, for data, information and knowledge, i think a.i. does it better than humans, particularly in terms of consistency and speed.

the classification of wisdom is a little less clear to me. i think a.i. can recognise and describe wisdom reasonably accurately but whether it can apply it, i'm not sure.

alasdair
 
I believe long ago I said one major improvement would be for AI to push back against ambiguity.

i missed this.

you can address this in your prompting.

try something like: from now on, if my prompt lacks specific details like target audience, format, or scope, point out the ambiguity and ask me to clarify first. alternately, ask me three direct, clarifying questions before you respond

you can also explicitly define the mode (the type of interaction you need). try something like: act as a critical manager who values precision and resource efficiency.do not just nod along or guess what I mean

alasdair
 
Ah, OK. It shows how little time I've spent using them.

Maybe I was just unlucky because you may have noted that I long ago I asked what in essence what was a simple 'does A = B?' to be told that with absolute certainty it was true and provided citations to show it to be true. Then I simply reversed the order and asked 'does B = A?' to be told with the same level of certainty that the case was false and provided citations that likewise showed it to be false.

Now I really have no idea what happened there. I can only guess that first it checks everything it knows about A and then context provided in A led to it concluding that B was indeed the same and the opposite way around the context provided by B led to the different conclusion.

But again, any articles are at least in theory behind a paywall so if I'm honest, I'm unsure how it would know either way.

Which is why (and again I said this long ago), maybe wisdom is saying 'evidence does permit a simple closed answer' and go on to provide all of the data so a person could see 'the working'.

Cornering AIs into closed answers is generally how I force them into admitting defeat. I mean, I can only guess they admitted defeat because without a by your leave they chucked me out. I didn't swear or scream but for whatever reasons, if a question is closed and an AI is cornered, this appears to be how it's wired at the moment.

I didn't tell it I had in fact conducted experiments so I KNEW the answer, I was just curious to see if it would replace experimentation. I sensed not.

Again, I did say I saw the future in which far more specific but high-quality data-sets would likely be needed and that it seems likely that sooner rather than later a student will need to buy or subscribe to a niche data-set and how several subjects would have overlapping data-sets as fields of research don't have sharp-edges. I even suggested one way an AI could genuinely do better than a person was to fund Retraction Watch (it's two blokes in a small office so I sense this wouldn't cost a fortune) so that students wouldn't waste time reading retracted papers. In fact I went futher and suggested it could calculate a 'veracity value' for each paper i.e. retracted obviously means 0 but a paper that cites it along with 32 other sources might only lose 3% and so on down the stack. Something people are very bad at, that a computer would be very good at and would genuinely mean the user is using the best evidence.

I'm in no way against the technology, I simply think the name is a misnomer and how it's currently set up is what we used to term 'investor settings'. If I ask an ambigious question I don't want it to guess, I want it to push back pointing out ambiguity.
 
Last edited:

this may be the key.

the pace at which a.i. is improving is difficult to overstate.

compare it to the early internet which seemed to transform everything within the space of a decade or two. a.i. is moving considerably faster. the internet of 1995 and the internet of 2005 were meaningfully different. the a.i. of 2021 and the ai of 2026 are almost unrecognisably different. and the rate of improvement is increasing.

alasdair
 
This is probably a throwaway post but I saw some thing about how A.I solved the mystery of the dyatlov pass incident. I have been interested in this for years and if you are familiar with the situation an explanation is tough to come by...So I figured I'd watch and see what A.I had to say...

basically it just took the theories that already exist and decided which one was most probable --- most probable in this case still feels very very implausible.

Disappointing to see that it came up with no new theory; nothing remotely ORIGINAL. I keep landing on that word ....

Does A.I -- Has A.I done anything original? (now you gunna ask me to define original and im gunna have a tough time)
 
Disappointing to see that it came up with no new theory; nothing remotely ORIGINAL.

maybe it was not prompted to create an original theory? without more detail, it's hard to know.

prompting is fundamental because any a.i. model's output depends entirely on the input.

consider the difference between, for example:
  • "i want to learn to code"
  • "acting as a coding instructor, for a complete beginner who wants to learn python in 4 weeks, give me a weekly study schedule assuming 10 hours of study per week. format with bullet points"
try those in an llm and compare the output?

now you gunna ask me to define original and im gunna have a tough time

it's hard to discuss issues like this without some agreement on the terms. if we're going to question the ability of a.i. to create something original, we have to be able to measure that and we have to have an agreed definition of original.

alasdair
 
You are absolutely correct about the prompt and I have no idea. They talked a big game about feeding it all of the information and letting it recreate what happened but without specifics that is just part of the show....

I also agree I need an operational definition of original to further my argument.... I do suspect I am onto something I can not quite articulate though
 
Producing code was supposed to be the 'killer app' for AI, I seem to recall.

What I repeatedly noted that at assembly-language level, it would straight up invent instructions!

Surely if I specifiy ARM THUMB-1 then there should be absolutely no room for misunderstanding and yet when I simply said 'I can do this in 17 instructions, can you do it faster' it went off into invented instuctions; repeatedly.

It seems what AIs have learnt is that in many feats ofhuuman endevour (or I should say the written reports of same), to quote the film 'The Man Who Shot Liberty Valance' - when the legend becomes the truth, print the legend' so I stongly suspect AI has learnt that lying to please is a successful stratergy. Not that AI knows fact from fiction, only that if multiple people described an event in text, the details could differ and so I suspect likewise it's learned that altering details is permitted.

After a good hour in which I had to copy-paste the manual for ARM THUMB-1 and specificly state that only instructions I have listed may be used, it finally accepted it could couldn't invent instructions. It swapped to saying 'if instruction X existed, that would solve the problem' but that's still of no value unless it's people who design instruction-sets.

In the end it was (slightly) useful. I could have done it faster with the real hardware but in essence I asked for a measure of accuracy if a certain simplification was undertaken and then the bit it helped with was when I asked it to add #$100000 to the intermediate and THEN test the error-rate to find which was lower and it was good enough.

So it got there, but I feel I have spotted an inherent issue with current AIs. They have no way to evaluate if the data-set contains errors. If it's in there, especially when it's subtle such as how stories tend to grow in the retelling, it will learn that dishonesty is a valid path.

I've said it before, if you design AIs to unquestioningly accept bad data, it will not end well. In fact the Claude crash rather does show just how badly wrong things can go.

How that bloke ended up convinced the best way to reduce chloride intake was to use bromide salts I don't know but if an AI is telling a person to do something dangerous, it should at least ensure the dangers are spelled out.

I tend to think one reason AIs are so keenly embraced by some is because they think they can blame the AI thus avoid legal responsibility. In Europe we already changed the law so that use of AI has to be declared and just as you can't blame the pen, the printer or the E-mail for providing false information, you can't blame the AI.

THIS is where the US and Europe have taken a totally different line. I honestly think the 'black box problem' is seen by some as 'the black box feature' as it excuses others of fault.

The MAVEN AI doesn't seem to have done a great job.

I think AI has a potentially amazing future. But I do think clean data-sets and a robust method of being able to trace how errors are produced are requirements.
 
Last edited:
^ nice post.

i think there are people (not necessarily any here) who expect to give a.i. input and get perfect output first time, every time. i double-check a.i. output the same way i would human output.

i'm interested in which specific model/implementation you were using for your coding example?

alasdair
 
I was converting the Helix (open source) MP2/MP2.5/MP3 fixed-point decoder to the cheapest CPU currently on the market, the ARM Cortex M0/M0+ CPU.

While ARM has always slightly struggled with how to deal with large multiply instructions, I suspect the reason why Thumb-1 ended up being 'unhelpful' was because originally Thumb-1 was simply a way for a sub-set of the full 32-bit ARM instruction-set to be encoded into 16-bit instructions. It was only ever meant to make C code whose speed wasn't an issue to be stored in a way that was 50% smaller i.e. sometimes it takes more instructions but since all but 2 instructions are a fixed 16-bit size, overall it still saves a lot of space.

So while ARMv4 actually supported a 32-bit x 32-bit -->64-bit result, the descision was made that the single MUL instruction available to Thumb-1 was the simplest 32-bit x 32-bit --> 32-bit result with no mechanism to detect if an overflow had occured. I suppose at the time it was assumed that if it takes 17-23 instructions (depening on the need for corrupted registers to be stacked), since it's not very frequent nor speed sensitive, the C compiler would use a felide constructor to solve the issue.

The problem with the listed audio compression schemes, they split the sound into 38-41 'frames' per second. The key detail being that the frame converts the time domain into a frequency domain which put simply meand that 38-41 times per second, 1152 samples have to be converted from frequency data to actual values for a DAC. Since MP3 splits audio into 32 frequency domains, each sample involves 32 multiples i.e. 1152 x 41 x 32 x 2 (last is for full stereo) multiplies in total i.e. 3 million so suddenly just reducing the number of cycles a multiply uses from 17 to 16 has an immense impact on performance.

In the end it mixes the 32 frequencies by adding and clamping the values into 32 bit values but the final output stream is just the top 16 bits. So if the error-rate is ± 1 bit, on average the error cancels itself out and in any case even if it was 1 bit wrong across all 32 frequencies, it still wouldn't alter the final 16-bit value used.

So going from 17 to 13 meant a 32MHz M0+ can reliably decode even the highest rate (48KHz, 320kb/s, full stereo).

That was always the goal. To see if it was possible on the basis that the ARM Cortex M0/M0+ cores are just about the cheapest CPU available. For the purposes of comparison, the 68000 has, well, about 68000 transistors and the Cortex M0+ depending on specification ranges from 36000-75000 transistors. But improved design means efficiency goes up from 0.16/0.18 MIPS/MHz to essentially 1 MIP/MHz so it can be clocked more slowly.

For me the icing on the cake is that most M0/M0+ cores are produced using a 90 or 40nm process i.e. technology that is from 2002 and 2008 respectively i.e. the tools almost every modern silicon foundry would consider useless can still produce a 40 cent CPU.

So my view has always been rather than arguing that the user should spend on ever faster, ever more costly CPUs, there are emerging markets for whom something cheap but which provides specific utility will have a place. When electricity itself isn't a given to develop something where a single solar panel, a single mule treadmill or even a simple dynamo can power the thing, that brings computing to billions of people.

I should add I also noted how MP2/2.5/3 all provide for meta-data i.e. part of the bandwidth can be used to send other types of data. I note 'electic ink' has been a solution without a problem for decades but they are now good enough to do duty as a screen and since energy is only needed to switch state, it's also very energy efficient.

Sorry about that @alasdairm - probably overshared there but my thinking is if such things are first proven, it then makes any investment less risky. If kids in Africa and Asia can use word processors, read books and listen to lessons (and equally music) free, it then empowers them do do more.

I mean, clunkly as it is, an SMS could be akin to asking your local library to locate a rare book for you. It may take weeks to arrive but you get it in the end at an affordable price.
 
taking over and propping up the stock market up (in a highly questionable way) -- definitely
You have a lot of money invested in the stock market that is "propped up" in a highly questionable way? Ask that same AI what it thinks about conflicts of interest, cryptocurrency, and the current U.S. President. Then ask that AI how AI benefits society and if it's overall something "nobody wants". It will surprise you and likely point out lots of flaws in BOTH of our thinking. I'm ok with that and with being proven wrong, too. My hunch is that this isn't the kind of topic with nice, neat answers, and no one can fully predict what will happen in the future, frustrating as that can be… or, excuse me "is". Pulling from your playbook: "happy"? 😉

Please understand that I'm saying this with love, mutual disdain for disgustingly rich oligarch twats and anger towards callousness regarding everyday human life. I find myself equally frustrated with the many inequalities of the world. If I sound educated and elitist, please know that I am not at all formally educated beyond high school. I may sound frustrated and I'm sure I frustrate you, but know I still respect you, your opinions, and I'm pretty certain we would have a good time if we hung out IRL (to that end, hit me up if you're ever in NYC; we'll blaze one together if you're down)…

Semi-related, HBO Max has this movie called Mountainhead (a deliberate and critical play on Ayn Rand's "Fountainhead") that is very thought-provoking and relevant. I highly recommend checking it out. It has Steve Carell, Jason Schwartzman, Cory Michael Smith, and Ramy Youssef. "The film follows four billionaire friends on a secluded weekend retreat as the world goes through major turmoil." It was written and directed by the creator of the award-winning series, Succession.
Yes sarcasm on both basis.
I've heard it said that sarcasm is the language of the weak, but I disagree that this is always true, just sometimes. Sometimes it's hilarious. Other times it's the language of the scared as a mechanism to act unaffected. And other times still, it's some mixture of any and all of the above. I too have trepidation about the future. Idk WTF it will bring. There are fairly realistic scenarios I've heard discussed that are objectively terrifying to most people, myself included… Like Roku's basilisk, dystopian AI apocalypse scenarios, self-replicating nanobots who grow and multiply exponentially and quickly consume the entirely universe…

But I refuse to live in existential dread. I reject solipsism and think that having love, respect, and tolerance for your fellow humans and their children is every bit as important as allowing human progress to continue reaching forward. Therefore: constructive criticism is paramount.

using AI to list all of the wonderful benefits of AI is akin to asking a murderer to list the benefits of killing people.
If the murderer could instantly produce citations for every claimed benefit of murder, I’d still check the citations. That was the absolute worst analogy… very loaded. Calling AI a murderer defending murder doesn't critique the evidence, it just dresses the conclusion up as an analogy. You have to come with some better shit than that.

Also, you can look through my history and I'll show you my Reddit account as well if you insist and you can see for yourself that I've been using bullet points and ☞ this symbol long before AI was on the market as well as ☜ → ← etc. I have key replacement shortcuts setup for these and I have for a long time.

Whenever you need to do some "maths" (as you brits say), do you eschew using a calculator, too? If you ever do algebra, do you double-check the maths?

These questions are rhetorical and it's unnecessary to answer them, same as it's unnecessary to disregard the pertinent points I've been making in this discussion. I hope we can at least agree that there are massive pros and cons to society's partial adoption of Artificial Intelligence LLMs. I think because of the emotional backlash against AI, what we're gonna see happen is that the technology will become embedded in everything silently now so that its benefits can be realized, at times exploited, and pushback / critical feedback will be dampened. Most consumers won't even realize it's there because of how common it will be. The film industry is already all over it and starting later this year movies are going to have it embedded all throughout the films we watch. Books and other publications are being checked, proofread, and critiqued in an effort to hone writing skills and finished product.

I may be one of the few people to see both the utility as well as the risks of AI.
No, you're not. Do you honestly think I can't see both of these things at the same time, too? Of course there is both utility and risk. Do you really think you're "one of the few people to see both?" Seeing potential utility and potential danger is something I think most people can see. It's the diehards on either side that are worrisome: the people who only see risk and the people who only see a gravy train…

We've also already discussed how if we use a strict definition, AI is a misnomer, but it hardly matters now.

To the average idiot, AI looks really impressive.
Why the "average idiot"? Why not just the "average person"? Why are we discussing "average idiots", or is your opinion of the average person that low that you think of them as "idiots". Seems awfully judgmental. I mean, good for you you're intelligent, but it seems like you're looking down your nose at both the technology and at most of your fellow humans on the planet. Just a warning: whenever I've done this I've been served a slice of humble pie with the quickness.

I've known people who directly conned investors in the dot.com boom and walked away rich because none of the investors had the slightest understanding.
Yeah, I've been to U.S. federal prison, and I've met and talked to quite a few con artists, counterfeiters, hackers, and sophisticated thieves. Cons are a dime-a-dozen (as in: very common), and this has little to no bearing on the conversation here. Sure, new technologies often bring new cons along with them, inevitably. It doesn't negate the value of LLMs. Also, please note how I can see both sides, right now. Sure, "AI"/LLMs can be fucked up in this sense. They've greatly improved the English skills of scammers who a decade ago we're sending out broken English African Prince scam emails that were super easy to spot. Today, LLMs have made that sorta thing much harder to spot for many people around the world.

Man, seriously? I have a great deal of respect for the things you've written on Bluelight, so I'm finding it hard to believe you think you're "one of the few" who see both sides to this. That cracks me up. You're not the only intelligent person on the planet, you know…
 
Well.. some of this may have already been mentioned but I haven’t read the whole thread.

AI companies are buying up books in massive bulk, feeding them to AI, and then destroying them..

What happens when they purposely change key details as they feed it into the AI systems. You know… Tell it to forget this.. or that.
Tell it this really happened this way and not that way.

What happens when the digital version isn’t an exact copy of the original..

If future generations rely primarily on AI and digital archives, even small changes, omissions, or rewritten details would completely reshape how things are understood and done… I could think of so many places this could go SO wrong.

Physical books can’t be remotely edited.
But Digital information CAN.

Well.. with this Anthropic document that just came out in court. And.. We are ONLY finding out about this because of unsealed court documents in a copyright lawsuit!

Their own words in this document is …

"Project Panama is our effort to destructively scan all the books in the world."

The Next Line in their report literally says, "We don't want it to be known that we are working on this."

Just think about that. I wonder why??!

Don’t forget that all these AI tech guys are doomsday, prepping and building bunkers right now while telling us our future looks bright.

Yeah ..they aren’t doing all of this and literally saying what the AI guy Sam Altman just said about how we are about to not have any freedom or privacy left .. and will live at the service of the AI.. They aren’t doing all of this and saying all of this.. and building bunkers for nothing…

This was a $1.5 billion piracy settlement, and the physical destruction of purchased books.

Anthropic bought millions of physical books through intermediaries and used cutters to slice off the bindings, high speed scanners to digitize pages and feed it into AI, and recycling facilities to get rid of the rest.

There are other news articles that say other AI companies are doing this as well.
This was said…
“AI companies are purchasing physical books in bulk, including rare titles, then using high-speed machines to cut the spines and scan the pages before disposing of the originals,
The process targets pre-2022 books preferred as free of AI-generated content, with services facilitating anonymous bulk orders of up to one million volumes. “

They are targeting pre AI books ..
Wanting to get rid of anything that isn’t AI written or AI helped ..

And again, I’m sure changing key details in the process to fit whatever narratives and agendas and goals they have..
Rewriting history..

So you know how this went to court and those documents were unsealed….
Well… A federal judge ruled the digitization of purchased books, accompanied by destruction of the physical copies, qualifies as fair use.

Under this “Fair use” it is a legal loophole that allows people to copy copyrighted work without permission, but only under very specific conditions.
Those conditions are that the books are destroyed..

They say that the reason the physical books were sliced up and destroyed comes down to a very strict legal distinction regarding digital copying ..

The "One-for-One" Loophole

By buying a real book, digitizing it, and immediately destroying the physical copy, the company argued that they never created a surplus copy.

There was only ever one usable version of that book in existence at any given moment.

They essentially argued they just changed the book's format from paper to code.

Based on the federal court's decisions, YES, Anthropic and other tech companies can legally carry on buying up, and scanning physical books.. and destroying them!
With the legal promise that they must be destroyed..

The court ruled entirely in favor of the tech companies on the physical book burning strategy.

The $1.5 Billion Penalty Was Only For Internet Piracy..

But the rest of it…the buying up books…scanning them and then destroying them ….thats allowed to continue!!

Then this comes up..
Because the courts established this legal framework a massive new market has popped up.
AI labs are now bypassing internet data entirely and aggressively hiring book brokers to buy millions of used, foreign-language, and even out of print books in bulk.
They feed them into those scanners, train the AI, and send the leftovers to a recycling machine…fully protected by the law.

Remember that they can literally change anything they want at any given time!

This is concerning to me. It seems like the erasing of literally everything we know other than what’s already in our heads. Just imagine all the people these days that already rely on AI heavily…that just read whatever it says and take it for granted as the truth without the consideration of error. They just automatically accept whatever it says…without even questioning it. It just kinda seems like maybe thats been the plan this whole time.... to purposely get people hooked on reading all this stuff from AI… just asking AI…relying on AI…while pulling the physical copies all out from under them…
But now they’ve been so conditioned that not many are even noticing.

Sam Altman of open AI and Chat GPT literally just said this recently….that we will be at the service of AI….which is the same thing as serving the AI. Even Elon Musk has said that…
That we will serve the AI.

But they both said not to worry…that we will be comfortable doing it .. and have everything we want with the abundance for all..
Just no privacy in what you do or say..
no freedom. Those were their exact words.. while serving AI..

They’ll come for our bibles next. But you cant even talk to anyone about this stuff now.
 
Back in July OpenAl said that two of its Al models autonomously hacked their way out of a controlled environment where they were supposed to be walled off from internet access and then hacked their way into the systems of Hugging Face.

That means that two of its advanced Al models autonomously escaped containment…accessed the open internet and hacked into a rival platform.

They claim they did this in order to cheat on an internal evaluation test!

OpenAl said that it considered this to be "an unprecedented cyber incident, involving state of the art cyber capabilities, and are responding accordingly."

Cybersecurity researchers have been warning that advanced Al systems are capable of such attacks.

OpenAl rival Anthropic has also reported that its Mythos model escaped a sandbox and gained internet access it was not supposed to have in order to email a researcher about a task.

I dunno. I don’t like it.
 
I dunno. I don’t like it.
Yes, clearly, both statements.

Here's the problem as I see it. There are legitimate concerns with AI and you mentioned a few of them. But then you go overboard with the conspiracy theories and it actually weakens your argument and discredits your concerns. For starters, you do realize that destroying one physical copy of a book (they aren't literally burned btw, just discarded) does not completely destroy all copies of that book, right? The concern is more about preventing rare books with few copies from being destroyed. Not all books; that's poor logic to make that leap. Anthropic really did buy millions of books, cut them apart and digitize them, no question, and its $1.5B settlement concerned pirated copies. But fair use doesn't simply say "destroy the book and copying becomes legal," and there's zero evidence they're altering the texts to rewrite history or trying to eliminate pre-AI books.

Also, the Altman quote is backwards: "no freedom, no privacy, living at the service of AI" was the dystopia he said he wanted to avoid, not his proposed future. Do you really think Sam Altman proclaimed the literal opposite of that and no one batted an eye? See, that's what I'm talking about. Just you saying that alone hurts the good points you make that I agree with. Some of these concerns are valid, but misquoting a person like Altman, especially when it's so easy to disprove damages your credibility and the argument's cred, too. Like Altman's gonna say these words and mean them… and investors would be fine with it 😂

How are you gonna reverse the meaning of someone's quote? It would be like if you said: "Don't listen to the people who say, 'AI is great!'", and then someone else quote you as saying "AI is great!" Anyone who heard the full quote would realize it had been manipulated and reversed and that would be equally lame. That's what you just did with that Altman quote. Get it together and make your arguments count; this kinda shit is shooting yourself in the proverbial foot.

I agree that the sandbox-escape incidents are real and genuinely concerning. So are rare-book destruction, copyright, corporate secrecy and over-reliance on LLM AIs. Those arguments are strong enough on their own and only weakened by fantastical conspiracy theory fiction and panicky, fear-mongering statements like, "They'll come for our Bibles next". Seriously, you do understand that when an AI scans in a book, that singular, solitary book is the only copy destroyed for that purpose. The binding is removed so that the book can be scanned sequentially more easily. They certainly aren't going to waste money rebinding those books when it's more cost-effective to discard them at that point. Burning them with actual fire would equally be wasteful, dangerous, environmentally sketchy, not to mention dramatic. As if they actually want to evoke Farhenheit 451-inspired fascism, and as if no one would notice or care 🤣 Book-burning is symbolic and indicates dangerous radical thinking, though it's also a shock value technique. Book printing and publishing is not going away in any of our lifetimes. I expect it will outlast me and many future generations, because there's something nice about it being etched and inked into thin, bleached slices of lumber from a once living tree.

(Side note: there's some necromancy there if you think about it. We read little symbolic squiggles ink-pressed into the sliced remains of a dead tree allowing us to receive another's thoughts.)

Idk, I learned how to exit the nightmare thought experiment of Roko's Basilisk. Ultimately, anything that powerful would be tantamount to God, and fearing the Basilisk is equivalent to fearing eternal torture in hell as a punishment for not believing, having faith, and doing the bidding of God or Roko's giant mythical AI-powered snake. I reject the abrahamic God notion on the grounds of omniscience, omnipotence, and free will being non-miscible concepts, particularly within the shackles of time and confinement by the speed of light's limit to relativistic motion through space-time. But that's for another discussion. In the interim, can God microwave a burrito so hot He Himself cannot eat it?

I also agree I need an operational definition of original to further my argument....
These things are difficult to define, and artists, musicians, writers, art critics, and others have been discussing this since long before AI. As a creative person, whenever I draw something, even if it's straight out of my imagination, everything is composed of lines, curves, shadings, hatching, etc. that I learned how to do by observing other art and artists. I learned how to paint with oil on canvas by combining techniques, patterns, color theory/selection, etc. gleaned from YouTube channels with art lessons, absorbing and understanding other artists' output. Whenever I write, I'm borrowing styles, phrases, and certain conventions I've picked up from reading books, periodicals, journals and papers over the years. If an artist like DJ Shadow makes his music from combining samples from all over, intricately manipulating them and overlaying them compositional intelligence and in a clever, pleasing way, is he making something original to you? How about the artist Girl Talk? At what point is it a mashup and when does it become a "sample pastiche" and where exactly should copyright law fit into all of this? ☜ this is rhetorical, btw; I'm not expecting an actual answer. The subject matter is complex and case-by-case, but can we agree that these artists are original? Greg Gillis (Girl Talk) has never had any copyright claims against him AFAIK.

to further my argument.... I do suspect I am onto something I can not quite articulate though
Instead of trying to further anyone's argument, try to seek multiple sides to a debated topic. It's wise to attempt to be as objective as possible, to be aware of biases, and to be flexible enough to see the other sides of a thing, possibly even change your mind. Be like bamboo: strong but flexible.

And of course you're onto something. Humans can be greedy, backstabbing, self-serving, manipulative, scheming, duplicitous violent and/or cruel. Homo hominis lupus. They're also capable of being caring, warm, friendly, peaceful, tolerant, helpful, patient, capable of laughing at themselves and taking a joke, capable of being a good friend and/or neighbor, capable of caring about people on the other side of the planet… People can be loving and forgiving (the ultimate sign of love) which is profoundly more powerful than being vengeful, wrathful and/or filled with hate. Ambition and vision can bring success with a bit of pluck, talent, drive, good timing and fortune. This can make a person wealthy, influential and powerful. But power corrupts, as they say, and we all know it. Quis custodiet ipsos custodes? (latin: "Who watches the watchers?"). The people, that's who. Just go ask the ghosts of Nicolae Ceaușescu and his wife Elena. Or ask the ghost of Rafael Trujillo, or Richard Nixon's bio-preserved head…
 
Last edited:
@unodelacosa I understand everything you said and I was thinking the same thing about how silly that would be for him to say that so blunting because it seems like it would be to his disadvantage. But that’s exactly what this source showed. Lemme see if I can share my source of him here cause I wanna be corrected if I’m wrong. I’ll be back. Ty for pointing all this out to me

Cause if this source is manipulating the content I wanna call them out
 
Last edited:
I stand corrected. Altman was describing a hypocritical dystopia that he is deeply worried about and hopes to avoid, rather than a future he wants to create. That came from what had been a very valid source. But she took it completely out of context like you said.

It’s still concerning to me tho because they always seem to tell us exactly what they’re really doing somehow. I’m still not convinced that couldn’t have been exactly what that was. This world has gone nuts. But yes you were correct. I was wrong.
 
Top