|
Gentleman Baller posted:Just checking, if open source developers genuinely don't want people to profit from their work without paying coders they could just publish their work under a licence that restricts commercial use, right? This is actually the current situation with LLaMA. It has a non-commercial license. Meta absolutely benefits from it, that's why they released it in a way that it was painfully obvious it would leak and then, once that happened, they made no attempts to stop it. And no one had any misconceptions about that fact, but the fact is we got a powerful foundation model and it's hugely benefited making local LLMs and their training accessible to normal people with a gaming laptop or something. Most open source models don't want to be non-commercial though. Most open source software is about contributing to society as a whole, being for really whatever you want to use it for, not just hobbyists. A whole lot of people benefit from Linux commercially, for example, but that's not necessarily a bad thing or take anything away from Linux. So, openlm-research is actually just on the verge of releasing their 7b version of a model trained in pretty much exactly the same way as LLaMA but without a non-commercial license, OpenLLaMA. Maybe they got tons secret funding from Microsoft and this is a 4d chess move to destroy society somehow. But if you believe their story they're just some students from UC Berkeley. And that's kind of a big deal, not too long ago training a 3b foundation model was a huge task that just took a lot. GPT-J wasn't even 7b. And now some students got above that just screwing around basically. And, as far as I know they're gonna keep going and train equivalents to the 13b, 33b, and hopefully 65b models too. As far as all the other pieces being non-commercial, people could do that but it wouldn't matter much. You don't really need much besides transformers and a few lines of python to make a model work. BrainDance fucked around with this message at 12:05 on Jun 3, 2023 |
#
?
Jun 3, 2023 11:59
|
|
|
|
| # ? May 20, 2024 16:04 |
|
|
This presidential election is going to be wild, and Trump is going to go all in on AI generated deep fakes. There was a fake of DeSantis released with his face and voice superimposed over the scene in The Office where Michael Scott wears a woman's suit. https://www.businessinsider.com/deepfake-video-ron-desantis-michael-scott-the-office-pantsuit-2023-5 https://twitter.com/AlexThomp/status/1662431349838413824 It's pretty sloppy at certain parts, especially around 30 seconds in, the video breaks completely back to Steve Carrell and his voice for a second in transition. This is pretty obviously a parody video but I don't think they are going to stop at noticeable parody. DeSantis as Michael Scott is a great comparison though, they look the same and have similar mannerisms.
|
|
#
?
Jun 3, 2023 13:57
|
|
|
SCheeseman posted:We're all captured, everything is by default. Copyleft licenses work from that assumption and try to thread a needle to create some kind of balance, but as they're a reactions to capitalism in a capitalist society they can't help but be compromised by it just like everything else is. GPL's commercial allowances are a part of this, a deliberate attempt to court capitalist organizations into contributing. Which they have, the Linux kernel is overwhelmingly maintained by people who are paid. This also means that volunteers have their work used by those entities, though since contributions make their way back it's a stretch to say that it's been taken. okay so what about the guy who invented the polio vaccine and then didn't patent it because he wanted everyone to be able to not get polio. is he an rear end in a top hat because capitalism exists, or what. i don't get what you are trying to say here
|
|
#
?
Jun 3, 2023 20:18
|
|
|
Gentleman Baller posted:Just checking, if open source developers genuinely don't want people to profit from their work without paying coders they could just publish their work under a licence that restricts commercial use, right? I'm not sure if this is the thread for going in depth over this? It's such a boring dry topic even if the ramifications are potentially huge. Not sure where else to talk about it though. I am planning on modifying CC-BY-NC-SA for use in my 'ethically sourced free range vegan public domain only' Stable Diffusion model. Creative Commons Attribution-NonCommercial-ShareAlike. There is a lot of existing argument over this not being a 'Free Cultural Work', I see it as what's known as a Clopen Set. Both Open and Closed. This is my line of thinking, I want this to become a Public Good. Free for all to use with the exception of commercial use of the model itself. I want no part and would make no claim over outputs of the model just the model itself. If you wanted to make something with it, slap it on a t-shirt, then sell it? All the more power to you, and all profits to you as well. Hell, I'd probably buy one myself. What I wouldn't want is a company charging for access to the model, for instance a website charging per image generated using the model, hiding it behind a paywall, requiring a paid membership to access, or more specifically Microsoft rolling it into base Windows itself. Unless they did follow the license and made Windows free? Not going to hold my breath there. The Adapt or Remix part is crucial since I am using the same open source training software as everyone else there would be no technical way of preventing it from being used in merges with other models. These merges are the current majority of the most popular public models. I wouldn't try to prevent this either I would actively encourage it. Attempting to climb to current levels of output quality from scratch by myself could take 3-4 years easily, where with even a modicum of extra help could easily be brought down to 6 months or less. That the initial outside merges would be 'contaminated' according to 'purist' thought of copyright infringement also wouldn't bother me in the slightest because it would have no bearing on my 'pure' public domain model or it's future changes. I would just keep adding new material and working on the quality until it matches then surpasses todays. This effectively eliminates the current 'ethical set' argument that irks me so much. AI Art isn't theft. Rent is theft. Here is another reason for the ShareAlike portion, there is already precedent. Dreamlike Diffusion has already released a modified noncommercial model, and it's already been used in several merges which also have to follow the noncommercial license. That this is a 'contaminated' model also doesn't bother me, no one is forcing you to use it. I know nothing replaces the advice of an actual lawyer but general discussions of are still useful for information gathering. There are aspects I don't know about such as, does the noncommercial part mean you can't host the model if you run advertisements on your website? There are bound to be avenues I haven't explored and I don't want to get this wrong and end up hurting people over it. This should be free and open for all 8,000,000,000+ of us to use which is why I'm taking these steps to see that it is.
|
|
#
?
Jun 4, 2023 00:13
|
|
|
Persiflagist posted:okay so what about the guy who invented the polio vaccine and then didn't patent it because he wanted everyone to be able to not get polio. is he an rear end in a top hat because capitalism exists, or what. i don't get what you are trying to say here Why are you acting as if the person you're responding to would think he was an rear end in a top hat? The point was that capitalism will exploit and try to capture even such acts of generosity (and they certainly have, see insulin, which also was released for free by its inventor to make is as widely available and cheap as possible), but that doing that those things is still worthwhile?
|
|
#
?
Jun 4, 2023 00:24
|
|
|
GlyphGryph posted:Why are you acting as if the person you're responding to would think he was an rear end in a top hat? The equivalent to this would actually he a permissive open source license too, wouldn't it? Which is what we're doing here. Which is what we're saying is a good thing. And the problem people seem to have with it is that open source contributes to everyone, including corporate interests. At least as far as this comparison can go, it's not gonna be an exact comparison no matter what because vaccines and AI models don't work the same way and aren't the same thing.
|
|
#
?
Jun 4, 2023 02:04
|
|
|
Persiflagist posted:okay so what about the guy who invented the polio vaccine and then didn't patent it because he wanted everyone to be able to not get polio. is he an rear end in a top hat because capitalism exists, or what. i don't get what you are trying to say here Yes a total, 100% rear end in a top hat. He should have been put to death for what he did. Exactly what I was saying. You are very smart. e: In the current regulatory environment it's probably a good idea to patent discoveries like that even if you want to distribute it freely, so other companies don't take it for themselves and fight attempts at establishing prior art. SCheeseman fucked around with this message at 04:15 on Jun 4, 2023 |
|
#
?
Jun 4, 2023 04:12
|
|
|
BrainDance posted:The equivalent to this would actually he a permissive open source license too, wouldn't it? I don't even know why it's a topic. AI generators are a fact of life now and like most other software there will be high-end commercial versions and free open source versions. What's important and matters is that everybody will have access to it. Corporations profiting from the volunteers that make the tools doesn't affect what the tools can do, who has access to them or the impact it will have on the art community.
|
|
#
?
Jun 4, 2023 06:19
|
|
|
SubG posted:... I have not mentioned EleutherAI, and don't know the status of EleutherAI. I am open source developer myself, so obviusly I am I support and defend open source. Edit: Theres nothing bad about open sourcing AI. Open Source algorithms is very good for society as a whole, including corporations and random guys and artists (despite artist favouring products pretty often). Theres nothing bad about scientist building corpus of data and distributing that with a free license. What is bad is when these "scientist" are jus a facade to build that corpus using copyrighted images. Tei fucked around with this message at 08:22 on Jun 4, 2023 |
|
#
?
Jun 4, 2023 08:17
|
|
|
Tei posted:I have not mentioned EleutherAI, and don't know the status of EleutherAI. Are you suggesting any specific open source developers are operating a facade for corporate interests or are you just speaking theoretically?
|
|
#
?
Jun 4, 2023 08:28
|
|
|
Tei posted:What is bad is when these "scientist" are jus a facade to build that corpus using copyrighted images. Sure, maybe. But all the evidence for that actually happening here in this conversation we're having about EleutherAI is some guy saying "yeah that's totally what's happening." For no reasons, that he'll share at least, at all other than I guess he just feels like it. We can list all the problems with open source software, but if it has nothing to do with the open source project we're talking about here then what does it have to do with anything? Why are we talking about it? The whole point of this conversation was in response to this StratGoatCom posted:The point is that these things are innately and inescapably megacorporate products and that the liberatory rhetoric is utterly empty, given precisely how they are made and what they are for. Which is just straight up untrue, no one has been able to produce any evidence of it being true about EleutherAI (and they're not even the only ones, I just posted about another group at UC Berkeley.) And saying "but it could be true and it would be bad if it were true" isnt making any point at all.
|
|
#
?
Jun 4, 2023 08:34
|
|
|
KwegiboHB posted:I'm not sure if this is the thread for going in depth over this? It's such a boring dry topic even if the ramifications are potentially huge. Not sure where else to talk about it though. not 2 get totally off topic talking out my rear end about IP law but -- as far as i knowzies -- a lot of the "noncommercial" clauses in licenses liek creative commons r basically untested and accordgingly dont have solid precedent. 1 example is Great Minds v. Fedex Office & Print Services, Inc. (2018) where a nonprofit sued fedex for printing NC licensed shizz for a fee, in which hte court ruled in favor of kinkos (or whatever teh fux theyre called now lol): quote:In this case, because FedEx acted as the mere agent of licensee school districts when it reproduced Great Minds’ materials, and because there was no dispute that the school districts themselves sought to use Great Minds’ materials for permissible purposes, FedEx’s activities did not breach the license or violate Great Minds’ copyright. Overall tho its not very well established what "non-commercial" means. wikimedia commons 4 example doesnt allow NC-licensed content based on some assload of reasonings, altho its up in the air whether u think they are smart or not. Personally i thinkz noncommercial licenses are cool 4 some stuff and bad 4 other stuff (ive licensed some shizz under GPL, some CC-NC, some straight CC, some CC0, etc). Anyway im with u if u are trying to devise a new license (altho i might read it and tell yuou that part of it is dumb lol) (USER WAS PUT ON PROBATION FOR THIS POST)
|
|
#
?
Jun 4, 2023 10:08
|
|
|
Persiflagist posted:i thinkz noncommercial licenses are cool 4 some stuff and bad 4 other stuff (ive licensed some shizz under GPL, some CC-NC, some straight CC, some CC0, etc). Anyway im with u if u are trying to devise a new license (altho i might read it and tell yuou that part of it is dumb lol) I'm not trying to be an rear end but why are you typing like this, here?
|
|
#
?
Jun 4, 2023 12:57
|
|
|
Tei posted:Theres nothing bad about open sourcing AI. Open Source algorithms is very good for society as a whole, including corporations and random guys and artists (despite artist favouring products pretty often). BrainDance posted:I'm not trying to be an rear end but why are you typing like this, here?
|
|
#
?
Jun 4, 2023 14:26
|
|
|
Corporations are going to try to minimize costs no matter how they can. They also don't care if art they receive is meaningful, the bulk of people want chintzy crap. Fine art will continue to exist for trained artists and the people who want it. AI art will probably evolve to the point where some AI art is considered fine art. Art isn't a profitable job, it never has been. I would agree that it should be, but the fact that it's hard to be an artist isn't AI's fault. I think that person just posts all over the forums and forgot which one they were in. It's not a big deal or worth a derail, it just looked a bit weird.
|
|
#
?
Jun 4, 2023 14:41
|
|
|
you are right that people want the chintz, but the fact of the matter is there once was an artist behind it. every ai generated book cover is a lost commission (excluding those who would just as well slap together something in MS paint, of which them using ai doesn't change much). supporting these models supports artists losing work
|
|
#
?
Jun 4, 2023 15:08
|
|
|
cumpantry posted:keep telling yourself this while corpos continue to favor ai and put artists out of jobs (artist defined as someone who doesn't use the kind of chintzy crap) Open sourcing AI doesn't keep it out of the hands of the corpos and the person you're responding to is actually in favour of generated art inheriting copyright from datasets. On the issues that actually matter to you, you are in agreement with them. Try reading the thread instead of busting in looking for targets.
|
|
#
?
Jun 4, 2023 15:08
|
|
|
i'm not arguing about the datasets
|
|
#
?
Jun 4, 2023 15:12
|
|
|
o wait i see what youre saying lol sorry (USER WAS PUT ON PROBATION FOR THIS POST)
|
|
#
?
Jun 4, 2023 15:14
|
|
|
The data sets are the works that AI generators build from. Given copyright would transfer, anything fed into them would have to be licensed, at which point your criticisms about it taking jobs away are roughly akin to criticising the licensing and use of stock art instead of hiring a photographer. If it's the automation itself that bothers you, any work using new techniques is taking work away from those who stuck to the old ways, not dissimilar to getting upset at electronic composers because they use digital instruments instead of hiring musicians.
|
|
#
?
Jun 4, 2023 15:23
|
|
|
cumpantry posted:you are right that people want the chintz, but the fact of the matter is there once was an artist behind it. every ai generated book cover is a lost commission (excluding those who would just as well slap together something in MS paint, of which them using ai doesn't change much). supporting these models supports artists losing work Every book cover made using software took jobs away from artists who used to do all the drawing on paper. Software is easier and cheaper and faster. It is also less *~artistic~*. Supporting software means supporting the loss of work. And yet, somehow, with every advance in automation the End of Work never shows up. Instead people end up working more than ever, just with different tools. I've been paying a lot of attention to AI lately and I have yet to see how this is inherently different. Faster, yes, and in an unexpected direction, yes. It still seems just another form of automation.
|
|
#
?
Jun 4, 2023 15:58
|
|
|
https://technomancers.ai/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/quote:Japan Goes All In: Copyright Doesn’t Apply To AI Training I have no idea what that blog is but they source https://go2senkyo.com/seijika/122181/posts/685617 which is in Japanese and I can't read Japanese but I assume it's a more reputable website. Anyway, looks like while all you guys have been sitting there arguing about copyrights and hypotheticals, the Japanese government has decided to pull off the bandages and let Japanese AI development go hog wild. Now your lovely fanfics and coverarts are fair game to ingest and use to train models to create fanfics and coverarts in your style. Now what?
|
|
#
?
Jun 4, 2023 15:59
|
|
|
My Japanese is still not good enough to fully translate a long form article like that, but it appears to be talking about existing law and specifically saying that they need to update the law to protect copyright holders. That article itself is also a summary of a Q&A session with a specific subcommittee - it is neither a direct quote, nor a stated policy goal. Hopefully the Google translate is good enough here (I cut out stuff related to overtime for teachers): quote:Kii Takashi Blog That original English blog was clearly being sensational for clicks. For Japanese news, I would look at NHK. If it's important/noteworthy, it will probably be there. KillHour fucked around with this message at 16:51 on Jun 4, 2023 |
|
#
?
Jun 4, 2023 16:45
|
|
|
^^ Can you read the minutes? I have no Japanese so i'm relying on google translate completely. I don't read Japanese either but this is the conversation that lead to this article is posted on KII Takashi's website, who is a member of the Japanese House of Representatives. He posted the minutes of the subcommittee meeting on his website, obviously in Japanese but the translation for the rational from google translate is quote:Minister of State Nagaoka I guess there is a specific section in Japanese copyright law that required enjoyment of the orginal material? That might be a bad translation and it means something else, but they don't think that AI meets the requirement to be considered used for copyright.
|
|
#
?
Jun 4, 2023 16:51
|
|
|
Luckily, a Japanese law firm wrote an English article breaking down the relevant part of the law. The whole thing isn't long so I encourage everyone to read it here https://storialaw.jp/en/service/bigdata/ai-14/14-2 Here's the opening. The rest is mostly an analysis of the implications: quote:As explained above, Article 30-4 of Japan’s Copyright Act, which allows a broad scope for the use of copyrighted works if it is for the purpose of “data analysis”, is said to be quite unique from a global perspective. Basically, the argument is that training a model is an act of data analysis, so it's explicitly allowed. It does on to say that since the resulting model doesn't contain the actual copyrighted data, it would likely (doesn't really give justification for this) be permissible to use the model as long as the output isn't infringing on its own. Here's the key point, IMO quote:Therefore, in my opinion, the act of generating AI software capable of expressing a particular author’s style by learning only the images of a specific author does not fall under the category of “unreasonably prejudicing the interests of the copyright owner” under the proviso of Article 30-4. I think that's where a lot of people already are ITT, so it's not that crazy. gurragadon posted:^^ Can you read the minutes? I have no Japanese so i'm relying on google translate completely. Sorry, I'm phone posting at the moment but it's pretty clearly talking about the law as it exists today and not some policy goals. I might try to translate it later as practice but no promises since it's a lot of work. KillHour fucked around with this message at 17:07 on Jun 4, 2023 |
|
#
?
Jun 4, 2023 16:59
|
|
|
No that's ok, that analysis that you posted was what I wanted to get out of a translation either way, so I'll check out that blog. Looks like Japanese copyright law is just written with more leeway. This is the law which I copied from that link Killhour posted.quote:This provision is Article 30-4, Item (ii) of [Japan’s] Copyright Act, which was adopted by the Act Partially Amending the Copyright Act of 2018. This does have wondering if other countries are going to follow Japanese law or try to maintain stricter copyright like the EU is trying to do. If this tech really ends up being revolutionary than the EU would be left behind.
|
|
#
?
Jun 4, 2023 17:20
|
|
|
Persiflagist posted:not 2 get totally off topic talking out my rear end about IP law but -- as far as i knowzies -- a lot of the "noncommercial" clauses in licenses liek creative commons r basically untested and accordgingly dont have solid precedent. 1 example is Great Minds v. Fedex Office & Print Services, Inc. (2018) where a nonprofit sued fedex for printing NC licensed shizz for a fee, in which hte court ruled in favor of kinkos (or whatever teh fux theyre called now lol): Thank you for that link. I personally think that outputs should be seperate from the model or dataset that created it, wikipedia couldn't host the model itself but should have no problems with any of the pictures made from it. I still have some time before release and will have to find an actual lawyer to talk this over with. Not looking forward to that. SCheeseman posted:generated art inheriting copyright from datasets. This in general I did not think of because of the copyright offices stance on output. Their guidance letter doesn't mean settled law though so I'll have to give this some thought. Thanks. Boris Galerkin posted:Anyway, looks like while all you guys have been sitting there arguing about copyrights and hypotheticals, the Japanese government has decided to pull off the bandages and let Japanese AI development go hog wild. Now your lovely fanfics and coverarts are fair game to ingest and use to train models to create fanfics and coverarts in your style. https://timesofjp.com/society/copyrighted-works-get-flimsy-protection-from-ai-under-japanese-law/ quote:The group also raised other issues they considered problematic, including the fact that Article 30-4 of the Copyright Law, which permits the use of a copyrighted work for machine learning, does not include procedures for gaining permission in advance from copyright holders. If I put out an article "US reaffirms requirement of seatbelt usage while driving" I don't think I'd get quite the same number of outrage clicks. It could be worth talking about their old law, but it was an old law, nothing actually changed.
|
|
#
?
Jun 4, 2023 17:26
|
|
|
cumpantry posted:keep telling yourself this while corpos continue to favor ai and put artists out of jobs (artist defined as someone who doesn't use the kind of chintzy crap) Using copyrighted images is a choice. You can make a worse bot with only images in the public domain or a license that allow derivative copies / use withouth attribution. Of course, using the best images available not matter the license give you a better bot, so is what they did.
|
|
#
?
Jun 4, 2023 18:08
|
|
|
Boris Galerkin posted:https://technomancers.ai/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/ Please be a bit less hostile toward your fellow AI cowboys.
|
|
#
?
Jun 5, 2023 02:56
|
|
|
Boris Galerkin posted:https://technomancers.ai/japan-goes-all-in-copyright-doesnt-apply-to-ai-training/ meanwhile:https://pc.watch.impress.co.jp/docs/news/1506018.html quote:The document titled "Regarding the relationship between AI and copyright" released by the Agency for Cultural Affairs and the Cabinet Office on May 30 clarifies the views on the relationship between learning by generative AI and the relationship between copyrights of products and existing works. was done. I would not try that interpretation in the legal environment that spawned Nintendo; open season on scraping would be repudiating Berne, which destroys their cultural industry. 'Research' just means that, research, not outright commercial use - bog standard Berne exemption.
|
|
#
?
Jun 6, 2023 04:34
|
|
|
https://www.youtube.com/watch?v=_MrGNlXRi9M Jim Keller, whose work exists in just about every piece of computing that exists on the planet, on a bunch of topics related to AI (and RISC-V). He never seemed much of a hype man, while he does have a vested interest in the technology he's pretty dry and low energy here. Starts rather business focused, mostly questions on corporate structure, logistics and the like. The questions, or rather the answers start getting existential near the end, Keller seems to draw a line between human intelligence and the work being done in AI which was surprising. He's pretty convincing, but I've been sold on that train of thought for a while. It does bring to mind the criticism I've seen brought up where asking GPT models to rationalize past behaviors only ever results in post-hoc narratives and therefore aren't "real" insight, when Keller posits that humans do the same thing. Also some typical engineer brained libertarian anti-regulation ranting, of course. SCheeseman fucked around with this message at 19:33 on Jun 13, 2023 |
|
#
?
Jun 13, 2023 19:30
|
|
|
I have read more than once Roger Penrose "The Emperor New Mind" book, because It was fun and full of interesting ideas. I did not agreed with the ideas in the book. To me it was all hand waving "our brain is special", withouth evidence. Sometimes reading a reactionary is fun.
|
|
#
?
Jun 13, 2023 22:17
|
|
|
SCheeseman posted:https://www.youtube.com/watch?v=_MrGNlXRi9M Not remotely surprising to anyone with the slightest idea of the difference between a jumped up autocorrect and an actual GPAI. Corvids are orders of magnitude beyond current 'AI', much less actual human level thought from a machine.
|
|
#
?
Jun 13, 2023 22:23
|
|
|
Drawing a line meaning they're making the comparison and describing similarities. Surprising because I didn't really expect an engineer as accomplished as Keller to get into the existential stuff, most of what he's written and the few talks I've watched didn't have him musing about those kinds of things. The second paragraph you wrote is a non sequitur. Who are you responding to?
|
|
#
?
Jun 13, 2023 22:50
|
|
|
Tei posted:I have read more than once Roger Penrose "The Emperor New Mind" book, because It was fun and full of interesting ideas. I've read his take and I agree with you. You'd think a physicist would make a stronger argument, but oh well.
|
|
#
?
Jun 13, 2023 23:12
|
|
|
So setting aside the art and chat gpt for a bit. Here’s a real general AI (again not LLM not image generator, actual AI) question. If the cybernetics assertion of consciousness fundamentally being a feed back loop ( or a feed forward outputting into a feedback) is correct, then the creation of a general intelligence AI is the creation of general purpose universal control.
|
|
#
?
Jun 30, 2023 22:08
|
|
|
Bar Ran Dun posted:So setting aside the art and chat gpt for a bit. Here’s a real general AI (again not LLM not image generator, actual AI) question. You might want to explain in more detail what you mean here: "universal control" isn't at all clear.
|
|
#
?
Jun 30, 2023 23:40
|
|
|
The box that says controller in the diagram in the link below of a control. https://www.eecs.umich.edu/courses/eecs373.w05/lecture/control.html You’d still need the equivalent of a transducer for whatever system.
|
|
#
?
Jun 30, 2023 23:46
|
|
|
Bar Ran Dun posted:So setting aside the art and chat gpt for a bit. Here’s a real general AI (again not LLM not image generator, actual AI) question.
|
|
#
?
Jun 30, 2023 23:53
|
|
|
|
| # ? May 20, 2024 16:04 |
|
|
SubG posted:If the question is whether the last statement is true, it is not, unless you're crafting your definitions such that it is tautologically true. "General intelligence" in the abstract doesn't entail the capacity to solve any specific problem or class of problems, nor does it imply that such a solution must exist. Here’s a general example before we had controls for burners in a boiler there was a person (a general intelligence) who would monitor the feedback from the boiler and would adjust the burners by changing burners / tips/ pressures. Before the pneumatic control valve was invented a person had to do the same thing in steam systems by manually adjusting a globe valve. In rocketry you had astronauts manually controlling burn to land on the moon. A general intelligence (a person) can get into a car and drive it on the road. That is to say being an (imperfect) universal controller is a characteristic of of our general intelligence. I don’t think it’s an extraordinary claim that an artificial general intelligence would be as well. Bar Ran Dun fucked around with this message at 00:13 on Jul 1, 2023 |
|
#
?
Jul 1, 2023 00:10
|
|