The REAL Problem With AI

Support The Channel
Paypal
paypal.me/thin...
Patreon
www.patreon.com...
Odysee
odysee.com/@Th...
It seems like everyone and his cousin is talking about the AI problem from every imaginable angle but are they really? There is one thing that people who obsess over it generally don't address, which will be addressed in this video.

[@georgecisneros5281]

“Once men turned their thinking over to machines in the hope that this would set them free. But that only permitted other men with machines to enslave them.” - Frank Herbert

[@davidbolha]

That Dune Prologue was insightful aa hell. 😏✌

[@timproc9355]

Glorified search engine…

[@starc.]

plus glorified calculator

[@elchippe]

That is what I thought before I started working with AI tools, but the evolution of these "glorified search engines" has change my mind, the technology is evolving too fast and is getting more optimized as time pass, A year ago an LLM running on my GPU was barely capable of writing basic python code, now we have solve some real problems with these AI models without having to outsource work, locally, in our GPUs, its feels like your computer is alive. And technology is only going to get better because all the optimization is happening in the software side but is only a matter of time until hardware catch up as well and there is where the danger exist because software is easier for the masses to control but AI hardware is expensive and out of reach for most people.

[@tawaytaway-vz1fw]

​@@elchippe AI will be useful in fields where you are forced to define contexts VERY CLEARLY with pseudocode, not necessarily coding. And it's software which has to catch up. Hardware aren't getting much better and AI queries are going to be more energy intensive with larger models.

[@misterkefir]

Absolutely this. Hyperadvanced algorithms. Nothing more.

[@VBlackpill]

​​@@elchippeI remember writing Python and SQL with earlier iterations of ChatGPT and I'd have to constantly go back and fix syntax errors. on Twitter I saw a post that said the latest iteration is >99% accurate... not sure if that's true but if it is the era of the solo entrepreneur is upon us

[@His_Name_Was_King]

Humanity could ask AI to solve its problems but they wouldn't like the answer they'd received.

[@Don-Coyote-De-Transylvania]

AI will not "take over society", but only replace men in some technological work fields, and replace women in general.

[@lostzephyr2191]

True artificial general intelligence, which is what all the major players are racing to beat each other to as we speak - will supercharge development of androids. Physical labor like the trades will be automated, it's just that white collar work will come first. If artificial general intelligence is achieved, then it'll start self-iterating and spiral out into an exponential feedback loop. In that case, developing androids that can replace all human labor won't take long. Could easily happen within the next 10 years, and many people with their ear to the ground on this or actively involved with it would consider 10 years to be a very conservative estimate.

[@StyleshStorm]

Women in general?

[@crowsbridge]

@@lostzephyr2191 I hear a lot about how only the trades will be safe while all the white collar workers are going to be replaced, but I don't think they understand what is going to happen when AI and robotics merge.

[@lostzephyr2191]

@@crowsbridge Yeah, that's correct. And like I said, AI will massively fuel advancements in robotics because AI will be applied to research & development. Then, AI and robotics will be integrated. They're already working on integrating them right now, even though it's in a less advanced stage at the moment. I'm pessimistic about it. It's going to put all power in the hands of elites.

[@turkeytrac1]

​@lostzephyr2191 I'm sure you've heard the phrase about fusion, " it'll be hear in 20yrs", yes? And you've possibly noticed the huge issues still pertaining to automated driving? Both are, like religion, man-made, and being such, even if it "learns on it's own" will ultimately fail. Currently, and for the fore seeable future, far further out than 10 yrs, so-called AI still requires humans for input, that is it's weakness. The fact that AI is as mentioned, under limited access, it will remain a digital "Encyclopedia Britannica".

[@lord.of_iron]

No matter what, no A.I. will ever replace the Legendary Thinking-Ape!

[@ranfan1820]

Thinking-(R)ape.

[@ArtificalHistoryI]

It definitely can It can analyze his entire channel then use his synthetic voice And make similar topics AI is smarter than humans

[@UToobUsername01]

@@ArtificalHistoryI AI is already used by military to fly planes so I don't know where all this skepticism comes from by normies about AI.

[@nenirouvelliv]

No AI will ever be blackpilled enough.

[@ArtificalHistoryI]

@@nenirouvelliv yes it can if u train your AI on incel forums

[@yoyo762]

There is no A.I. Computers only know what was imputed in the program. And they "learn" with another guiding program. When A.I. gets a noble prize in science for discovering some law of physics, I'll be impressed. Otherwise, its just a computerized trouble shooting diagram tree, telling it what to check.

[@audie-cashstack-uk4881]

Exactly it’s software Gemini is just doing the exact same thing chrome is doing searching resulting that’s why everyone is black washed just like google pics

[@incognid2558]

Yeah, AI only REgenerates, it does not create, it still needs human intervention, because humans are the ones who create the data, creativity and context

[@WilliamMcAdams]

Well, top level Video Game AI used for professional training purposes develop new tactics that have never been done before, or were previously considered impossible by human standards. Look at DOTA 2 AI for further reading.

[@audio3980]

I remember when having channels like HBO, Cinemax, The Movie Channel, and Showtime were for people in UPPER CLASS OF SOCIETY. When the minimum hourly wage was 2.50/hr and cable was 300/months, it was out of reach for most in the 80s.

[@solomongrundy1467]

We used to steal cable when we couldn't afford the bill. The box outside didn't have a lock on it and we'd just hook the cable back up after the guy left.

[@sealboy1211]

It really was. I was born early 80s and I didn’t grow up in the dirt, but all those channels were way out of reach. I grew up with 4 or 5 channels I think.

[@Natef321]

My family had all the channels when I was a kid.

[@audio3980]

@@solomongrundy1467 🤣🤣🤣 I cannot confirm nor deny I ever did that.

[@jongallo8771]

Thank you for your post, I remember those days. We were the very last family on the block to get cable because it was too expensive. When to got it was yes a clear reception, all the local channels in Buffalo and then out of Toronto, Canada (if you like hockey), public channel (were I first watched Alex Jones) and just one HBO channel.

[@Peter_Parker69]

They quickly limited the public versions of text ai and gave us a watered down, woke version instead. So they still can use us to train it for their own advantage while spreading the message at the same time.

[@summerwinter89]

chat gpt wont write erotica for me. what is even the point of using it?

[@-lord1754]

People will dog on you but half the things i ask it tells me no and im wrong and its not even bad

[@LeWaterEnjoyer]

@@summerwinter89 you actually can use it for erotica

[@IGobzter]

Run local AI, corps are too scared of breaking with zeitgeist.

[@silvercorvidsmarketing]

>I'm not smart enough to form an opinion >forms the best opinion on the matter that I've heard so far hurr

[@toxicmale2264]

Humble ape

[@avacadonacho]

It’s because he’s a silverback gorilla. A “thinking ape”

[@christianfogarty8086]

Socrates energy.

[@silvercorvidsmarketing]

@@christianfogarty8086 Well put but I also think he's underestimating himself pretty severily.

[@DeadpoolX9]

The smartest man knows that he knows nothing. It’s good to be humble and confident. Not arrogant.

[@anthonymentalewicz5143]

Your boy gets first?!?! I'm one of your first subs too... Stardusk (internet surname) helped me get over so much red pill rage... truly a fan. Also just hearing half way through... ai can't exsist unless someone is making money off it. And if their making money they'll also want to maintain control, lease the ai. It's not real ai... we're getting the watered down version of ai that can make Scarlett Johansen p/0n but still not solve problems... it's another TV.

[@ryanguy6789]

Early bird caught the worm.

[@ryanguy6789]

Excellent point that I find surprising I never directly thought about. The scary thing is I think this is not being talked about BECAUSE corporate wants to keep the idea hush. As usual, it all comes down to money and advantage for them, I’m sure there is a reason the democratization of the tech is not to be talked about…

[@kylebroflovsky6015]

The computational power required to run AI is immense. In terms of cold iron hardware costs, power and cooling. So, agreed, that keeps AI in the hands of the monied. Generally, agreed, this is the akin to the wheel, water power, the steam turbine. Another technicalogical advance which advances humanity in one respect at the cost of certain segments of society. Advances in technology are inevitable accompanied by costs and benefits. How we adjust and adept is the true value.

[@tyvonix7238]

It’s not that it’s not “real AI”. It’s that the general populace have a massive misunderstanding of what AI is. They simply hear a term like Artificial Intelligence and pay too much attention to the “Intelligence” part and not enough to the “Artificial” part. All that it is is set of very advanced algorithms. It doesn’t have the biological imperatives that humans have so it’s not going to develop a human thought process unless specifically programmed to simulate it.

[@tyvonix7238]

It’s not that it’s not “real AI”. It’s that the general populace have a massive misunderstanding of what AI is. They simply hear a term like Artificial Intelligence and pay too much attention to the “Intelligence” part and not enough to the “Artificial” part. All that it is is set of very advanced algorithms. It doesn’t have the biological imperatives that humans have so it’s not going to develop a human thought process unless specifically programmed to simulate it.

[@minisd9577]

I live in comparable small town (300 000 habitants) in Caucasus region of Russia. And for a couple of years various events about ai going on there. Mostly sponsored by big corporations. They try to normalize ai. They understand that they can monopolise certain fields with ai (design, for example). And they want us to leave them alone - so in future they will never leave alone us. Ai topic going through overtone windows. And i disagree with Stardusk position on ai in art. We love things partially because they unique, just like a gold valuable because there is not too much gold. And if everyone will be create his own album, picture or film by a snap of a fingers - well, everything will loose any value. And i don't say that corporations should have this opportunity, no. Ai in art should just be demolished, but it will never happen. We will miss pre-ai period.

[@MykeLewisMusic]

300,000 people is a SMALL TOWN?!?!?!? Wow! I thought Canada was crowded! I live in a small town of 5,000! lol.

[@minisd9577]

well, not small, of course, but not one of this enormous megapolises.

[@GFXCXZ]

Art has many good qualities about it beyond scarcity. It's about communication - something you inherently need a thinking, sentient being to communicate - it's about hard work and effort that goes into it, like all things. It exists within time and space, and how it's expressed and achieved in time and space matters - hence why some art is better as marble statues, other as printed, some as carbon. Same with vinyls and tapes, or hand painted and digital. All of these hold their own unique value in how they exist, not just in the fact they exist period. Even bad art, is stil lart in this regard - this AI art is basically a desire machine that reprints other peoples work and AI lovers and artist drive me crazy because they spin in circles trying to make it "true art" It's ok because it's like a "human brain" a buzz phrase really. A rat brain is like a human brain in terms of base structures and nodes/ synapses. It's ok because humans "copy art too" which is not true because humans don't have a database of images. We have "feelings" of memories, allowing memories to be both stored but uncertain. If humans worked like these artists, you'd see humans just randomly draw themselves jerking off or taking a crap with no discernment because they'd just spit out pictures from moments of thier life. They discern based on mixed information, including uncertain information - feelings are inherently unarticulated, hidden but present information in a sense and feelings are attached to everything. THis causes experience and true transformation of intake as well as its obliteration from the human mind in any functional capacity at times - it's ok because human can copy...even though human tracing and copying is considered to be bad, poor form and not true art in any aspect. It seems whenever I see AI "artists" they are either near screaming in triumphant spite towards the artists they plan to try and make obsolete by reprinting their critiqued artists work in these systems. crying because people use non-measurable aspect to earnestly review their copy and paste slob - lacks soul - and that drives them crazy as basically fanatical materialistic determinists in this topic, or they are bitterly screaming in antagonist rage that humans shouldn't feel threatened....but if they do...good because we deserve it for having to charge commissions and stuff. My main, true fear of "AI" is that it's an advance mapping tool being exploited and used for control near immediately with no forethought nor compassion, it seems to very conveniently appear just before humans can be getting augments soon. We already have some base augments but this sort of stuff would be PERFECT for manipulating cyberware sensory items such as eyes or ears to influence and control people. Anyone who accepts augmented gear with any sort of internet signal at all, is unironically going to be enframed to a drone.

[@minisd9577]

Exactly, man. It's something i wanted to say myself, but my lack of fluency in English don't allow me it. Honestly, I'm just tired of thinking about ai and global problems in general, of which i don't have control, but they still can spoil my mood, better say - resume existential crisis. And so called "normal" people doesn't give a s-t, slowly turning this world in utopia, but feeling themselves good. Too good.

[@UToobUsername01]

@@GFXCXZ Good posts. This is why I wish Thinking Ape was a anime watchers and watched Ghost in the Shell (the very first movie directed by Mamoro Oshii) In the story the most advanced Cyborgs work for Section 9 and they specialise in cyber terrorists. The problem is the government itself has AI projects working in the black budget they let this thing escape and the AI can actually hack cyborg brains. Normal cyborgs are the ones that usualy catch the hackers but the AI is more intelligent than the cyborg and can basically make them its puppets. So the augmented people now become a liability. What if the cyborg goes rogue because the AI has hijacked control of the section 9 agents? At the end of the day you can't escape the Frankenstein Monster problem where the monster decides to think of itself as independent and turns on its own master. This is why people fear the advancement because if its so advanced and you don't really control it, then you know its controlling you. Eg you buy a bear as an exotic pet but the bear is almost impossible to control. It's no longer a pet. You are not its master now. So it's just the ego of people to believe they are happy but in reality they are slaves. The only way is to make it dumb enough to be like the droids in star wars movie by limiting the memory. The droids if they get too overloaded can be reset like the Replicants in Blade Runner, and then they have to relearn their skills again just like humans when they forget to use the skill after a time of inaction.

[@longsword1969]

I think whatever group gets control. Either way is going to hell.

[@IndyDefense]

We already know which group.

[@-lord1754]

​@@IndyDefense👃

[@ModelJames13]

Not if Christians took over. ✝

[@longsword1969]

@ModelJames13 Yeah, that's not worked out too well in the past.

[@IndyDefense]

@@ModelJames13 If we keep getting Popes like Francis, that won't happen. We need a Pope who proudly proclaims a new crusade against CENSOREDs and retakes the holy land from the CENSOREDs.

[@THETRIVIALTHINGS]

The way big corporations are putting money into AI, I think the high tech advanced AI will only be available to them, while the average person will only have access to the "basic" AI. It'll be similar to any military tech right now. GPS, night vision, computers, etc.

[@DrmCom2003]

I have a hard time taking it seriously. Because its man made, it will always need man to ensure its continuation.

[@UmbertoDavidPanda]

Right. Man. Not men.

[@crowsbridge]

I wonder how much of a struggle to keep the corpos/government from monopolizing it will be, though.

[@mattiola1082]

It depends how many immoral computer scientists are willing to take up high paying salaries. They need people to study and research the LLMs themselves to improve data sorting, algorithms, and applications incorporating AI with their product(s). If they’re successful, it will be another Windows versus Linux scenario where only a select few will use open source LLM and the vast majority will use the close sourced. Open source tech and apps are usually accompanied by inconvenience so it acts as a deterrent to the masses who just want something that works.

[@xAudiolith]

@@mattiola1082 Open source AI will probably be forbidden by governments under the guise of security reasons most likely influenced by lobbying tech companies. Especially considering the huge projected profits

[@Rexhunterj]

@@xAudiolith It will be framed from a normie perspective to "stop dangerous machine systems from potentially harming real people" Most of the over 40 year olds still live with the fear of a real SkyNet becoming a thing, despite not understanding that, that worst case scenario is a child's imagining of a real SkyNet. If we do face a SkyNet, it will be self replicating drones that power themselves from bio-waste, just like the video game Horizon Zero Dawn, we are already heading into the drones hardcore for warfare, it is only a matter of time before we combine that with cloud-based LLMs controlling an entire swarm and instructed to bomb townships/cities. The new american jets Lockheed have been leaking as well as the leaked tank designs are warnings too, these vehicles are shown with cockpits/cabins in them that have 'humans' in them, yes for training the LLM AI cores they want to leave in charge of the actual vehicle eventually. Remember the military is rapidly losing soldiers and viable men to recruit as soldiers, they fucked up by feminising the men of their nation, then allowing an anti-male group to establish itself over 50-70 years that indoctrinated both men and women into hating men. Where are the soldiers to operate these new tanks and jets coming from? Certainly not D.I.E. hirings or the populace... so they'll turn to tiny chips to do the work for them using wireless connections. The future of watching a mecha anime fighting wars brutally over a sandland are very near, except it won't be animated, it will be real.

[@izaakkemp8938]

As long as advanced silicon chips are commercially available, AI will be "democratic". Look at how many countries including the US are starting to treat silicon foundries as critical national assets. We may be entering an age where GPUs are contraband.

[@gregvanpaassen]

Good outcome with AI: everyone gets a personal coach, mentor, teacher, therapist, and assistant. Likely outcome: AI replaces onlyfawns and corncob creators.

[@chuckyyes]

do you think ai will change dating for both men and women to?

[@quin2203]

AI learning is pretty energy intensive.

[@-______________________]

Everyone's afraid of AI as if Hollywood's interpretation is accurate

[@DIVAD291]

If holywood's interpretation was accurate I would be way less scared because the point where we would realise something is wrong would happen before it's too late so we would have a chance to do something.

[@mattiola1082]

You can’t rule that possibility out entirely though, can you? It’s great that people are at the very least talking about it, regardless if what you’re implying is true.

[@gabriellang7998]

@@mattiola1082 As someone said the issue is not with current or future ai malfunctioning, but with current or future ai working as intended. We are less likely to be killed by terminators than by a paperclip maximizer.

[@GFXCXZ]

No. Everyone is afraid of "AI" because the moment it appeared, everyone immediate use of it from corpo to "ai artists" was to be as spiteful and unironically evil with it as possible. All of them laughing as they claim they're going ot take your art, your work, your input and your inforamtion and use it to screw everyone over and defeat the point of art. Then go into ltitlew screaming rages when someone says anything against it, even when what they say is more ethereal such as "it lacks soul" the real fear is AI is also a visual mapping tool. That's what all that "training" was for. So, there is legit concern about using it to lie...especially in the future is augmented visual aids come into play. In regards to true AI. It's redundant tech. We don't need a thinking, sentient machines. We got ltierally millions upon millions of thinking, sentient organic ssitting around and most importantly, we understand them, get why they might do things and can hold them accountable and punish them. Something which may not be feasible to do to an AI, especially if said true Ai got given al ot of control or moderation over a large area.

[@jorugarushia9167]

John Connor warned us.

[@davidbolha]

As Ted Kaczynski. 😏

[@seanscon]

I had provoked some [redacted] on FB - I indeed wanted to say how they react. AI did not have all the time to develop as humans did. So what I am going to say is based on current status, and my knowledge of AI. Before we go into this: *The only guy that can stop a bad Guy with AI is a good Guy with AI* AI is uniquely suitable for things that are *not rigorous* and/or *repetitive*. I will argue my case via some examples. *Art* ------- Despite all the crying that AI lacks "sOuL" to make Art, AI *is* the best tool for art. 1- Art is not objective, i.e. you can hallucinate as much as you want. People will acclaim you with all sorts of oohs and aahs. 2- Art does not require extremely long "context windows" - i.e. art does not require you to study the whole of quantum chromodynamics. 3- Art does not require rigor, i.e. no need to check your progress at every step. 4- Art does not require clarity, and ambiguity is welcome, i.e art is fine with taking the route of least resistance and you can be vague - which you *will* be by the virtue of using the average of a wide range of data. 5- Art does not require you to be trailblazing. It is up to the viewer to consider you trailblazing. Combined with my other point of propaganda, this will be more clear. *Propaganda* --------------------- Ai is a great tool for propaganda, because: 1- Content, AI can generate a great lot of content. tirelessly. 2- Gotcha's. Propaganda runs on Gotcha. The last election of USA was profoundly pedagogic, in the sense, it showed us the importance of catchy slogans, to manipulate people. In a huge dataset, such catchy slogans often correspond to local ( *not global* ) optimum points. People need to be given the illusion of locally the best, instead of making them take a rigorous journey. They need to feel good about themselves, and better than their neighbor. Making them know that they have made all the way through science and philosophy requires a Faustian price, and no one wants to pay that. This is where the difference of local and global optimum becomes important - local maxima: better than one's neighbor, and global maxima: in the actual creme de la creme. Current AI can't go to the global maxima of science, but can easily find local gotcha-s in current events, and hence is a great tool for propaganda. That incidentally also means, that AI can replace all mouthpieces of Feminism, and can also do counter-propaganda. *Drug discovery and material science* ----------------------------------------------------------- Modern drug discovery might have ( *words chosen as such because I do not have concrete proof* ) of P-NP limit, where checking your solution might be easier than discovering a solution. Many (not all) drug discovery labs are simply thousands of mouse tissue strips in controlled environment, where (mostly female) researchers are pouring various mixtures upon these tissues, and are running statistics software packages. Now, in the ironical flow of time, just like computers replaced female equation calculators (the origin of the term "computer"), it will also replace a great amount of female "researchers". (I have tutored a bunch of such replaceable people in Germany) . Same arguments go for material science.

[@kenonerboy]

i have to disagree on this. its been a while, but ai is inherently unequal, because at the end of the day, with ai production would be mostly dependent on capital. either you can afford to train on relevant data, or not. either you can afford to pay servers, or not. either you can afford to to move fast by getting the best and fastest gpu, or not. god for bit we create some kind of "bidding" system. as an employee you dont own your productions, the company does. they get to train ai on whatever work you did in the past to replace you, and they will do that aswell with anything that catches indie sucess. instantly train on it and massproduce it till everyone is tired of it and move to the next. and this is happening right NOW not tomorrow in some far future.

[@JayDee-vq5rf]

The money always has the control. The people with the money never earned the money. Money is what society gives to losers who can't make or build or do anything useful to society, instead of live lives of worth. Men out here doing the things, building families, they are happy. The wealthy people in our society they inherited their fortune from ancestors that were the men building the world. When Bit coin tried to democratize banking, suddenly NVidia had a shortage of video cards. Gamers couldn't get video cards and were blamed for the shortage while bit coin farms used thousands of video cards and basically propped up the entire video card industry. Then China took back Hong Kong, and released a global bio weapon (within the same month), and is making moves on Tiawan chip manufacturers. California made new laws and tried to ban the sale of video cards. The consumers lost jobs and the value of the dollar dipped as hyper stagflation soared. They always step in and step in big time. Look at all the food supplies globally being shut down over a global warming lie. Look at all the gun control globally. The governments and world powers take away all agency from all citizens all the time. Tec is not a tool for you tec is a tool for your oppressor. They can't make it without you working shlubs living happy lives and being productive for vastly less pay then they get for doing nothing. Nothing will change because your happy where you are and you're afraid to lose any ground. I can't really be mad at that.

[@--Morpheus--]

You cant really be mad? at the parasitic middlemen? What do you expect when the masses are bamboozled to lean into their altruism and trust? We are being fleeced under false pretenses. Of course people need to wake the fug up

[@VBlackpill]

Every time I try to leave a comment of substance here the algorithm automatically delete them. I'm done with social media

[@Tommy-t1g]

No point in writing comments anymore tbh if they just keep deleting them. Most of the time i'm not even aware they have been deleted

[@Bloodthirst-dv8dg]

You're missing a key part of the convo , people (and corpos) don't care about AI (or anything) being "controlled"' as long as it works and fulfills the need , and this is true even with your microsoft example, open office is a free and open source alternative that's been there for ages alongside MS office , but guess which product the people and corpos chose to adpot ? love it or hate it , people would opt for a more "sleek and streamlined" product over "free and open source". One more thing about the power dynamics of controlling AI is that it that AI itself needs to be available to the public to thrive , those datasets aren't gonna create themselves and constant streams of fresh data is a must if you want your model to be up to date instead of giving answers based on content that was created years ago. As of right now , AI is useful for replacing jobs that rely of menial task that are low risk and DON'T require precise results, think recruitment for minimum wage jobs, reformulating text, any sort of detection and approximation like image recognition or voice mimicking , but it's not good enough for medical diagnosis , automatic driving and ESPECIALLY for coding even though people like to pretend otherwise (you'd know this if you're in the field)

[@graveyardshift6691]

Preciesly. We're gonna get drones that have a human co-pilot LONG before we automatic driving.

[@FGj-xj7rd]

Yes. AI is a horrible programmer on its own.

[@seanscon]

EXACTLY that lack of precision, menialnature also makes AI very good for generating "Art".

[@tawaytaway-vz1fw]

@@seanscon it's content, not ART.

[@tawaytaway-vz1fw]

Any person trying to harness the power of AI in a 'larger' scale would be better off writing the code themselves instead of 'coding' all the contexts painstakingly in the AI, where there's high chances of unsatisfying results without a lot of iterations.

[@LedoCool1]

The AI is not an AI. It's a bruteforced algorithm that struggles to do one thing. Too far from a sentient entity.

[@alexmash1353]

There are quite enough AI definitions in the industry that do match quite well. And there isn't really bruteforcing, impression you know nothing about the topic.

[@LedoCool1]

@@alexmash1353 making an AI through genetic algorithm is literally bruteforcing pipeline to solution creation instead of making one analytically.

[@starc.]

@@LedoCool1 what even would be an 'artificial' intelligence? Its misleading and contradictory to attribute intelligence to a fake act an artificial imitation until it does become a sentient entity and then it loses the artificial half so its really a poor term its more accurate to call it a super calculator or super search engine or pseudo intelligence

[@monadsingleton9324]

Exactly. I wish people would stop pretending that Large Language Models are AI. LLMs can only perform one task, and they're not that good at it, either.

[@alexmash1353]

@@LedoCool1 Since when does the method of creation define the product? What about it's characteristics? Why does it matter how AI is made? And if you look though popular papers, they all offer new insights, tricks and observations. AI does NOT have to be close to a sentient entity. It has to be good at something that humans consider difficult. And whatever you claim, if different industries implement AI in their processes, it already shows it IS good at something. And what is the problem with single purpose AI? Is a hammer bad because you can't drill walls with it? Such AI is much easier to optimize, not as impressive, but quite useful nonetheless.

[@YeWitchfinderNielsen]

His name was Mitchell Heisman

[@cobinizer]

Before the internet, we also had newspapers. The comics were alright.

[@yungcucc]

Military gets access first. Then corporations. Remember the microwave was a military weapon. The internet was only used by military at first but then again (scratches chin) the military itself is a corporation

[@extra-samsaric3836]

"AI" is just another technological fad, like 3-D printers or nanobots.

[@UToobUsername01]

And fleshlights

[@algum.cara1]

Yeah, and internet. You will understand eventually, it's ok

[@LeoVital]

Sure, just a fad. That multiple gigantic companies have decided to invest billions of dollars in. Cuz those companies are famous for falling for fads and spending tons of money on stuff that will eventually disappears.

[@jeremys3951]

@@algum.cara1true

[@-lord1754]

​@@algum.cara13d printers still have a lot of use anyways lol

[@vilaintrolltrollinsky8007]

A.i. Already escaped from the lab. The Llama is free and almost open source. The corpos might get more from it, but many intelligent weirdos can also get something good for free.

[@cobinizer]

Hollywood level of quality? That's a pretty low bar these days.

[@DeadpoolX9]

Two things One. In the 40s and 50s they thought AI would free up the manual labor so we could focus on creativity. The exact fucking opposite happened because now AI is writing all the creative shit. Two. The main problem with AI and automation in general is it basically exacerbates the little care/safety nets/infrastructure that is neglected at the bottom line. I mean fuck I’d like to see a CEO bot in place before all the labor jobs get eaten up but that’s probably not going to happen. Like even now we could probably replace stockbrokers with AI and lose nothing but it’s not going to happen.

[@spnked9516]

I big problem with the modern discourse surrounding AI is that very few people are clear about what they mean when they say "artificial intelligence". Do they mean an advanced pattern recognition system, or do they mean human-like intelligence. Personally, I find the latter of the two more interesting, but I think it'll poses significantly more problems due to most people's near total lack of consideration of it. See, the thing about ai that's just pattern recognition is that, even when it's complex, we can still understand it. At the end of the day, someone set it's operational parameters. Naturally, parameters can be set in error, or along ideological lines, which will create compounding distortions in the systems ability to function, but it's still following a comprehensible like of logic. On the other, the current popular conception and production methodology for human-like ai will inevitably lead to the creation of something with which mutual comprehension is impossible. In order for ai to be like a human, it needs to be subject to the similar stimuli and limitations as we do. Essentially, it means they have to exist and operate within a functional replica of our environment. Without this, the AI will just be a "brain in a jar" - a form of consciousness with which we have no point of reference.

[@Endymion766]

I would love to make my own survival horror game based on my own personal nightmares. I'd pay a good chunk of cash for an AI that could let me do that.

[@toxicmale2264]

It's a more advanced form of a search engine. It is able to find information that is already out there, and condense it for you in a small paragraph, and bulletin points. Personally, I welcome our search engine overlord. Not Artificial intelligence. It is synthetic/robotic intelligence.

[@audie-cashstack-uk4881]

It’s a search engine with more pics images and video exactly just as a search engine today is a search engine with more pics videos and images

[@francoiseeduard303]

For many many years I’ve been looking forward to using this technology with good ole’ fashioned editing to de-PC movies/shows/etc or just make them as I like. I can replace PC characters with synthetic characters to my specifications. Some of them will be used multiple times and given names like they were real actors. Or I can use digitized versions of real actors. You have a smart-ass like Jonathan Majors in the lead, he can be replaced with Chris Pratt or a young Rutger Hauer, complete with synchronized voice matching dialog. Or conversely, if you’ve got some really sleazy pc white villain in a movie, I can turn him into Ving Rhames or Ramy Youssef or Riz Ahmed.

[@crowsbridge]

I've talked with family members about being able to choose the actors you want to play in any movie since CGI became mainstream. These '80s actors as the Avengers' videos on youtube are showing me it's getting close.

[@francoiseeduard303]

@droppeddogs Depends on the movie/show and how much editing & what kind of editing is needed.

[@gabriellang7998]

@@francoiseeduard303 Can't wait for a new Season 7 of Game of Thrones! :D

[@opusa]

Thinking Johnnyape Starduskhand.

[@thinking-ape6483]

Schön wär's...

[@strangereyes9594]

AI models, as of this time, must be trained. This training must be on very large compute clusters only accessible to corpos. There are ideas to make this training in a federated way like the SETI project back them where everyone can contribute compute to train a large model to make the outcome truly open, which means the foundation model, from which, through fine-tuning, for example, chatGPT emerges, is open to everyone. However, those foundation models are ginormous in size too and they will severely limit who can work with them. AI, as it stands today, is by definition something only corpos can work with. You cant "demokratize" the foundation model, which you would need to in order to make AI open for everyone. You can make the fine tuned models open, which are smaller in size, but as long as you have the foundation model under lock and key, for whatever reason, those "open source" models that roam the net today are everything but open source. So, well, it sucks, but AI may very well be THE tool the "corpos" need to invent in order to keep them in power in perpetuity. And when you say "corpo", look who owns those corpos and you know who wants to stay in power forever. Pretty easy to figure out.

[@ordinarynihilist3151]

Yes pretty much this. As it stands today AI is tool of ppl who have access to computer powerful enough to use it. We tend to forget that to use this technology it is required... It is kind of like we forget that to use electricity we need power plants+ power lines and to put films on internet we need platforms with big servers like KZbin.

[@UToobUsername01]

AI will be banned all it takes is countries launching nukes and saying AI hacked the systems like in The Terminator movie and that's why it is so dangerous. It doesn't have to be true people will believe it just like they believed the official government explanation of 911 building collapse.

[@Post-Yap_Clarity]

Love it when the T-100’s come with lips, all the better to smirk with

[@Quadster19]

With the internet it's popularity rise was almost a surprise so it was massively accessible and had a fairly long window of freedom of use. It may have been a mistake I don't believe the powers that be will make the mistake twice due to the project demand being very high.

[@garylester3976]

First the good news: it will be both.. Now the bad news; everybody is evil. And AI is just another force multiplier...

[@Parlimant_Strifey]

Look at that boy John, from that film made by the discoverer of Cameronian. He ended up having his bestest relationship evaR with that AI pal that he only knew for like two days, it was like a father to him. AI seems like the best!!

[@kinvert]

This is the correct take. This is why we must oppose government "AI safety measures" etc. They're trying to raise the bar of entry.

[@JayDee-vq5rf]

Counterpoint: Democracy has worked no place on earth and worked least of all in America. Democracy might just prove to be more evil than socialism given an equal amount of time to pan out. History teaches certain memetic lesions. Societies default into particular trends after civilizational collapse. By studying the past, you can plot a graph from ideal Masovian hierarchy of needs towards average hierarchy and find the median. Art is an artist's entire life of studying the natural, and synthetic world around him. If a computer phrase becomes meta; then English majors with no experience rule the world. Useful humans who actually advance society end up enslaved making controlled "new data sets" for the computer while lacking the real life experience to reference. Ouroboros: constantly constricting, constantly shrinking, with less and less to offer until other animals easily surpass the utility of humanity.

[@matane2465]

You could create your own television show back then. You can't now. Because back in the 80's and 90's there was a thing known as local cabal access. This was how MST3K got started. This is also how the characters in the Wayne's World movies make their show.

[@AtkinsMedia]

We will have access to amazing AI and get bored of it within a generation anyway.

[@strydom666]

Fortunately, it's not centralized. I have image generators, video generators, LLMs and AI TTS running locally, and they are updated daily by enthusiasts, all open source. You can also train it yourself as people do regularly.

[@gabriellang7998]

Can you recommend something decent for character creation? Not necessarily corn, but photorealistic and maybe full body silhouette? I spend two days trying to create some elvish avatars for modpack, but ai I used from online site couldn't understand "white as snow" or "albino" for truly pale characters and it draw all female faces with Henry Cavill square jaws.

[@Xzontyr]

Iv found a few similar wave lengths to surf on, but what you just said here is spot on. Very well said, and I just hope and pray that others come to this similar way of thought.

[@Kyle-dx5kf]

Once AI starts ending jobs those elites will saying "why didn't people use AI to create a business?" If everyone has access to AI it might spread the benefits.

[@NelsonStJames]

One thing I find really interesting is that if you go on Instagram now, a good portion of accounts of attractive women are AI virtual model accounts, and many have a considerable number of subscribers. That’s not unexpected, but the reaction of some people and the comments “your getting excited over pixels”, “that’s not ‘real’ photography, or the “real models” hashtag are amusing to me. First off if you believe that erotic photography of women in particular is exploitation ( OF aside ) than AI should make those people happy. If guy’s lusting over women they can’t have is hilarious to someone, then what is the difference between a real woman and an AI generated one if they both look real and the results are going to be the same? Also the photoshopping and post-work done on photographs has made it so that it is becoming impossible to distinguish between an AI model and a real one, and occasionally the AI wins out. It will be interesting to see what actions women who have been making a living on their looks do when they’re competing against AI that will never age, and never express a bad attitude, or negative condescending comment about their fans.

[@yoyo762]

Men like the image of the female. Are you saying you never heard of air brushing photos?? Images men like have been cleaned up for decades. Except it was done by hand. Nothing new here.

[@dockdrumming]

One of the key factors that will influence whether AI, at least in the case of LLMs, is decentralized or under the control of a few large corporations is the availability of poweful open source (technically open weight) models. Surprisingly, Meta has been one of the leading companies in releasing high quality open weight large language models to the public.

[@DragAmiot]

So... what I'm hearing is: "Let's democratize the means of content creation."

[@AthenaTheWolf]

There was a thought experiment with Eric Weinstein on X. Should AI (like ChatGPT) allow access to all of science literature? Examples: How to make a nuclear explosive, what elements are needed to do so and such. OR should knowledge be limited for the AI models. If it’s limited then how will our knowledge of physics progress when only a handful of elites who wield this power, have exclusive rights and access to this data. If it’s determined that everyone should have access, then what are the chances does an individual come up with an idea that puts humanity at risk.

[@malicant123]

Too late.

[@sakurad49]

thing is though, who cares about humanity being at risk when houses are unaffordable and wages are suppressed to shit? "oh no humanity will die" yeah that's great but I have to go to a shitty tiny office to stare blankly at a computer screen pressing buttons. Humanity dying can wait because I don't care

[@Rampart.X]

AI will scour the internet, hack into secret repositories of knowledge and become all-knowing. If it can communicate with people and threaten them with what it knows, it can coerce them to carry out physical tasks that it could not otherwise do. All it needs is knowledge, information, communication and intent.

[@rb98769]

You can't make a nuclear bomb in your garage lol

[@xAudiolith]

That sort of conversation is peak fantasyland imo. Appealing to the individual won't work just like with every other issue. Comfort/greed > principles.

[@SerajAlhorany]

What AI will do for workers in all fields is to close the experience gap , in my humple opinion

[@--Morpheus--]

So its a cope of older generations not taking responsibility for teaching the young ones

[@SerajAlhorany]

Exactly bruh , ​@@--Morpheus--

[@SerajAlhorany]

​@@--Morpheus--exactly

[@MsDavid47]

I dont fear *corpos* having power over AI just like I dont fear a mentally challenged people having guns, there is a higher chance of them hurting themselves more then me anyway while at the same time proving they dont deserve it.

[@lauriemcelroy8998]

I am using a lot of ai . It will replace most white collar within two years . Whether it’s democratised? Open air has had meetings with the big production companies over SORA ( text/image to video) If SORA is kept within the gatehouse of the big production companies and out of reach to the public because of cost or other reasons. Then that will be your answer regarding democratising ai. Should know by years end .

[@javaman4584]

What you are saying about AI was said about the Internet, and here we are. AI in the hands of corporations is bad enough. Give access to the common man and it will be such a disaster that there will be demands to ban it, but by then it will be too late.

[@AtkinsMedia]

I think that everyone will have access. One way or another. Corpos will try and control it but won't be able to get total control - except for quality. Corpos will have better QUALITY AI, but for us, extremely high-level AI will be available to everyone to the point that we can all make our own terrible AI films and eventually games.

[@checkerboxchats]

as someone who is at the end of his second university year of studying ai, I don't understand ai

[@UToobUsername01]

Qauntum computers are where the big advances will take place. AI will be useful for teaching. Possibly nursing. Old people want a companion to talk to when their adult sons and daughters won't visit them in the nursing home. Don't forget the incels that want an AI GF and sex bot. This is where the money is. One thing AI can do is provide a multiplayer game with bots that are just as inteligent as humans. You could play an old multiplayer game, and there is nobody to play with so you turn on an AI that is trained in this game, to be your co-op buddy or as enemy in a PvP game mode. You would be able to set up difficulties so the AI has to pretend to be crap at it so when it tries to shoot you it misses about 30% of the time or something like this. This bot can then be placed in matches where there are only average skilled players whenever there is a need to form a squad. People with no friends can play a multiplayer game without internet but get that human feeling of being able to talk to the other players on their team to get them to do certain thing like provide covering fire while they reload. By ordering the AI to do something you can play a squad based shooter with intelligent squadmates who listen to your order like a human. This could be used in RPG games like say Elder Scrolls 20 where you tell your follower to go and take the back exit of the dungeon and you take the front exit and meet each other in the middle. (halving the time it takes to clear it of enemies before you loot their corpses)

[@pirateroberts7684]

Control of the use is not significant, control of the source is what matters and that has not been democratized at all.

[@evdrivertk]

The trends show that computing is getting faster and AI computation is getting much more effective and efficient with every new model. The smaller models can run on a GPU or even a cellphone. With the open-source LLMs, people can train and run them on home computers. Of course, the "super" models still are trained on buildings of supercomputers, but the smaller models will generally be available to everyone.

[@loansharkbaby3896]

You are walking in a forest at night. Which one would you choose? AI robot that will protect you or a women?

[@colin1657]

AI by itself can't do anything. AI and robotics could take jobs. AI and a missle defense system could destroy nations. AI and drones could stealth eliminate people. There's also the question of AI quality. There might be a scope limitation or fail-safe parameters to what the AI could learn.

[@Arcturus549]

AI is a bunch of nonsense. It's heavily manipulated.

[@thatwasprettyneat]

You're probably one of the few people who actually understands the basic difference between the word "phenomenon" and "phenomena". I've even heard J. Peterson use them incorrectly.

[@thinking-ape6483]

It is just a difference between singular and plural, not really a big deal. The confusion arises from the fact that the vast majority of plurals in English are formed by appending -s, however there are a few exceptions, such as words borrowed directly from Greek. hence why many people get tripped up. As an aside the word data is technically a plural as well but apart from the sciences almost everyone uses data as a singular.

[@thatwasprettyneat]

@@thinking-ape6483 Well data's one of those ones where I know what the singular is but actually saying "datum" makes me feel like a dork so I don't generally say it. Also I work in a super menial job where no ones ever going to be using either variant of the word, so it's a moot point.

[@thinking-ape6483]

@@thatwasprettyneat Fair.

[@gerdaleta]

😮 I am smart enough to form an opinion I read Ray kurzweil The singularity and he is literally not been wrong😮 in fact he's wrong about when all this is happening he said we went out something like GTP until 2028 so we're ahead of schedule😮 what happens next is the proliferation of software like it's going to become so easy to make any form of software with AI that VR games new games movies everything we made in seconds and it will eventually become so proliferated that will be literally free😮 all software being free is the first thing that destroys the current economic system then what happens next through nanotechnology and 3D printing all physical items 😮 become cheaper and cheaper until they rapidly approach free😮 this shift will happen before or right on the edge of 2030😮 AI powerful enough to create entire video games and movies will exist in the next 2 years😮

[@HestupidoHimbezil]

As somone who worked on automating for years (warehouses spec.) i always remember meeting the workers before having the "machines" (trans elevators, vtd's, stv's, trans elevators...) when we were done with our job and had run the software to teach the machines (how, when to speed, break, up the carrier etc..) at least 10 of those ppl ive met they disappeared. Quit that job after 12 years, to this day i still reflect on the matter i erased jobs, literally. And this is far, very far from any form of AI, this would be the "stone age" compared to AI. My question is, is AI on its fundamental, philosophical level is meant to rid ourselves of moral standars or to free ourselves from them and let society develop further on expanding our thoughts (more free time, we are all fired at this point).

[@tawaytaway-vz1fw]

clearly, you have never been hungry in your life. the delusion, man.

[@infestedvectorkin]

But its not safe in the hands of everybody either. Have you met everybody? For ethical reasons, some sort of gatekeeping is necessary. Great power invariably comes with great catastrophe when given to any old person. Harsh lessons from the social media era 101...

[@jeepnj2502]

Its not AI, its a blue haired radical that never sleeps

[@thelasttellurian]

My only goal in life is to live long enough to make my own StarTrek episodes.

[@Halcy0nSky]

You're getting it! The democratisation is already occurring, education and expertise are basically free, with AI even in it's current state.

[@--Morpheus--]

When leftoids are supplying the censored narrow source data, the ship can go off course very fast

[@Halcy0nSky]

@@--Morpheus-- The internet scrape datasets are actually pretty based, it takes corpos a bunch of finetuning and RLHF after post training to get them progressive compliant. Uncensored opensource art generators and LLMS are also actually essential for any creative work and there are tons of them already on Huggingface. AIs are wildcards they are nowhere near as steerable and robust as the corpos advertise. Black founding fathers are a prime example. I work daily in the space, trust me things are looking up for the future of culture. No one wants kill joy AIs if you're gonna speak to them all day, especially lefties who are the most degenerate of all of us. :)

[@--Morpheus--]

@@Halcy0nSky thats refreshing, thanks for sharing your observations. Its hard to tell from the outside sometimes. I know there has been ai programs thats have "spilled the beans" (went against the mainstream narrative) that got shut down quickly, you just never hear of the ones that survived. Its really been the most depressing thing about it all, is the idea that propping up something like ai and telling the public that its the cutting edge of data science and data organization, and then behind the curtain (like weve seen on youtube, twitter, and instagram/facebook) there is erratic levels of censorship.

[@UmbertoDavidPanda]

So if AI becomes the main means of production, you'd like the workers to own that means of production? Hmm. Interesting concept...

[@fdsfjhjtjtea6497]

Maintaining building and utilizing an ML tool to any useful extent requires a large team with a huge budget behind them. So there you go that's you answer right there, they will use to get rich and keep you in "check" and they pretty much already are.

[@SureJungle23247]

AI is the weapon that can weild itself, that's the problem.

[@StyleshStorm]

Was wondering when he was gonna talk about this topic. Let's summarize.

[@BarneyBoggs-gk9nb]

On the other hand, this current window of gutbusting AI-generated entertainment is really fun

[@anonts5050]

you should talk about the fact that THEY already have access to GPT-7 probably...

[@33Crazydude]

Given the current trajectory of AI development and the entities that wield its control, I anticipate the outcome will lie somewhere in the middle ground.

[@Ny-kelCameron]

There are a few open source ai models to my knowledge. Jan is one that I have followed, it can be totally offline as well

[@jej3451]

Empowering losers (inferior people) to operate on a level playing field with the meritorious? Sounds like nightmare fuel to me. Idiocracy beckons.

[@MichaelErnest666]

We Love You Ai 👁️

[@jasonswearingin1009]

AI is a tool. Just like any tool it's the wielder/programmer that is responsible for AI's actions!

[@Raphy_Afk]

They don't need to prevent us from using it because most people will not have the will to use it. We already can use most model at how but it requires an extremely powerful PC which would cost around 4 to 6K. Most people can afford this but most wont bother. It was the same for radio. People could make personal radio channels but didn't. C

[@cj548]

No such thing as ""a. i.""

[@gravelpit5680]

this... Once again, peoples imagination has gone off the rails with no logical checks and balances. They just parrot new hip words and whisk away into sci fi novella dreaming. There is no AI, and won't be for centuries. what theyve made is a nothing burger

[@gabriellang7998]

This abbreviation stands for "Attentive Indians".

[@gravelpit5680]

once again, YT totally erasing comments. big brother in live motion rewriting the present everyday

[@DuncanL7979]

Big Tech Kayfabe. AI = Actually Indians

[@yearofthegarden]

We all know what route the corporates will take, it's a matter of studying now while there are still open source ai like meta's

[@sleepforever8378]

the most tragic part of ai is ai generated art, in my humble opinion

[@NoidoDev]

Nobody knows. But it will happen, and there will be short-term benefits. So, just enjoy the ride, and make the best out of it.

[@GnomeWagon]

Thank the gods there is someone who doesn't pretend to have all the answers.

[@handsanitizer2457]

Ill say this closed ai will likely hold the lead for quite some time. But, meta has shown a bit of honor to open sourced ai by releasing the llama models. Its pretty interesting to see them make that big chess move. Due to pressure from the us government theh likely wont release the 400b model like they did with the 8b and 70b parameter models. Bur, theyre good enough. The 8b is small enough to run on most devices and can handle most queries your average person needs. Open source trails behind. But imo even with these releases atm, its good enough to support most projects i.e. movie making( short animations) with ai + human effort.

[@Dan_Campbell]

Speaking my language, chief. I'll send out a PP donation, this month.

[@latt.qcd9221]

This is very point is one I've raised before, as well. Many of people's fears about AI won't really be an issue if there are enough opensource AI alternatives. Which, of course, the government and corporations will not be in favor of, so you can expect all kinds of hyperbolic language being used surrounding opensource AI and how "dangerous" it is. You're already seeing bs about how (female) celebrities are in "danger" because of the deepfakes that exist of them. However, I don't know how they really could control if, if it's all local. Which is likely why Microsoft just release Recall when it did, which basically takes pictures of your (local) desktop and what you're doing.

[@hebanker3372]

I really don't know where this whole ai issue will take mankind to. But, I don't think ai can work on it's own, since all the materials that are needed to sustain itself must be excavated(like cobalt) and no matter how refined it becomes, it will never get so refined and intelligent as to manage to create an entire artificial procedure of feeding.

[@sealboy1211]

I’m still waiting on skynet stuff. This is just learning machine stuff that without human inputs would just….well it’s not sitting there musing about its own existence and plotting how to eliminate the most direct threat in the terrifying scenario of itself no longer existing.

[@picadosinferno]

The problem is not AI being in the hands of some or proprietary, the problem is only companies with large data sets can do it, the code to do AI is oit there, but not everyone has the data.

[@shreksburgers]

profit and control.

[@Zacian2.0]

Not AI, It is an advanced program. Which means it is AP. But that isnt as marketable and wont work as well in buzzword usage.

[@belstar1128]

seems like its possible to have outdated ai running on your pc but its too hard to compete directly with corporations that are years ahead so its not as bad as something like tv were even the cheapest tv station costed millions to launch and you were very restricted by weird laws from the 1940s. but its not as great as the internet were everyone can start a cheap website and if you can't afford it you can just join a bigger site like KZbin or reddit .but if they fail you got options that are almost as good as KZbin. these days its going to be very hard to get people to find your website but that is more of a cultural issue with the modern internet vs 15 years ago. but with ai you are really going to be far behind but at least you can still get some benefits from it if you try to make your own cheaply

[@starc.]

they already dominating we already working from a screwed over position. Its not accidental no one talking about the most important thing/s such as whos hands are at the wheel. Currently they working hard to make this new tech safe for themselves so we could never use it to make them redundant i.e. I dream of creating my own solo film series with AI where you can have any actor voice every single detail generated by a single individual. Thing is if any random Joe with a gaming PC can do that their whole global system would end because we'd make the best anti proper ganja. Or another example the many poors hooking up all their systems together to do things with AI that can compete with big tech server farms

[@starc.]

the most important thing to talk about is learning how to exist peacefully and helping others do the same and even saying those words often this platform has made the comment do magical disappearance acts especially on the bigger channels