[HN Gopher] Don Knuth's MIP, 64 years later
___________________________________________________________________
Don Knuth's MIP, 64 years later
Author : todsacerdoti
Score : 71 points
Date : 2024-05-25 03:57 UTC (19 hours ago)
(HTM) web link (nathanbrixius.wordpress.com)
(TXT) w3m dump (nathanbrixius.wordpress.com)
| hyperhello wrote:
| " I asked ChatGPT to render an image of Gomory and Knuth
| pensively sitting atop a mainframe. It refused for privacy
| reasons. Here is the best it could do:"
|
| Why? This took me right out of the article. No one needs this
| energy.
| chx wrote:
| Completely agreed. Instantly closed the tab without looking at
| the article when I saw the distinctly AI rendered image.
| awesomerob wrote:
| Shame, if you'd kept reading you would've been enlightened by
| the author's shocking discovery that modern computers can
| solve problems orders of magnitude faster than computers from
| the 60s or 90s. Just mind-blowing stuff, who'da think it
| jgalt212 wrote:
| really? I find AI art perhaps the only non-obnoxious use
| case.
| lxgr wrote:
| I find the idea of "AI art" ok, but something about Dall-E
| 3 specifically just riles me up.
|
| It seems like OpenAI went out of their way to make
| everything generated look a certain way, while that's
| clearly not a limitation of GANs in general.
|
| It makes every single blog look like they have the same
| illustrator, which I really dislike.
| robertlagrant wrote:
| So many non-AI article illustrations are just Alegria
| anyway.
| jgalt212 wrote:
| I call it Corporate Memphis 2.0
| falcor84 wrote:
| I don't know if anyone "needs" this energy, but I liked it.
| Just as I did when I was 6, I enjoy having illustrations when
| I'm reading, and am generally ok with them being AI-generated;
| it's better than using random stock photos.
| sedatk wrote:
| AI generated imagery is usually imbued with inaccuracies and
| errors. Not to mention the ethical problems of using artwork
| without consent for training the AI.
|
| So, it might be worse than a random stock photo in those
| aspects.
| glimshe wrote:
| I'll let the courts decide the legality of the artwork, I
| can't myself see any ethical problems. I agree with the
| parent that an AI-generated image is a lot better than a
| stock one and enhances the material as long as a human is
| able to validate that the image is appropriate for the
| context.
|
| In this case, I'm glad that the image was included. I was
| about to close the tab but the image caught my attention,
| and I kept going.
| chx wrote:
| I do not trust the corrupt courts of the United States to
| decide on anything. Or, more charitably even if the
| blatant exploitation of artists can't be proven to breach
| any existing statute -- and the law is always behind --
| even then I will shorten this obvious truism as "all AI
| art is stolen art".
| FeepingCreature wrote:
| Why not go one step further? All art is stolen art.
| thfuran wrote:
| Because that makes no damn sense.
| FeepingCreature wrote:
| There's literally a saying that explicitly spells out
| "great artists steal." Art is imitation plus adaptation,
| and everybody used to know this, artists included, until
| they all spontaneously forgot about a year ago, around
| the time when the artificial imitation machines started
| becoming commercial threats.
| seanhunter wrote:
| Yes. That saying probably[1] started life with the quote
| from TS Elliot, who said
|
| "Immature poets imitate; mature poets steal; bad poets
| deface what they take, and good poets make it into
| something better, or at least something different."
|
| The composer Igor Stravinsky "took" that and changed it
| to "A good composer does not imitate; he steals." and
| after that quote was further changed in the LaTex
| documentation (believe it or not), Steve Jobs took it and
| changed it to be Pablo Picasso saying "Lesser artists
| copy, great artists steal".
|
| It doesn't seem there's any evidence of Picasso saying it
| although Steve Jobs said a lot that Picasso said it.
|
| [1] https://quoteinvestigator.com/2013/03/06/artists-
| steal/
| richard_todd wrote:
| It's amusing that the sequence of events in that story
| makes a great example of prior art getting regurgitated
| and altered by humans. If he were an AI bot, we would say
| Steve Jobs "hallucinated" a fact about Picasso.
| chx wrote:
| Yeah. Still there is a human mind and effort between the
| original and the copy or we wouldn't care that much.
| That's the fundamental difference.
|
| Also https://news.ycombinator.com/item?id=40475364
| FeepingCreature wrote:
| Sure so call it dehumanized art, but don't call it stolen
| art just because that sounds more severe.
|
| It's just the '24 version of "You wouldn't download a
| car."
| Max-q wrote:
| If you go to art school, what you do is study and analyze
| what artists have done before you. We humans learn much
| in the same way as an AI. We study what has been made by
| others.
|
| I think accusing AI generated work to theft, is a based
| on the wrong assumptions.
|
| It's not just art. What do you do in computer science
| courses? Or math, language, physics, or any other
| subject? You study what has already been made. What you
| produce based on that knowledge is not considered theft.
|
| Have you ever seen an AI image that, if a human made it,
| would be considered a theft?
| thfuran wrote:
| >If you go to art school, what you do is study and
| analyze what artists have done before you.
|
| You don't have to go to art school to produce art. Even
| if you do go, there's nothing stopping you from producing
| art unlike what you studied.
|
| >We humans learn much in the same way as an AI. We study
| what has been made by others. I think accusing AI
| generated work to theft, is a based on the wrong
| assumptions.
|
| What assumptions?
|
| >It's not just art. What do you do in computer science
| courses? Or math, language, physics, or any other
| subject? You study what has already been made.
|
| Most of "what has already been made" is not "copyrighted
| work for which you have no license or exception"
|
| >What you produce based on that knowledge is not
| considered theft.
|
| It certainly can be.
|
| >Have you ever seen an AI image that, if a human made it,
| would be considered a theft?
|
| I've certainly seen AI produce things that would be
| considered copyright violation or misappropriation of
| likeness or similar offences if they were to be used
| commercially.
| FeepingCreature wrote:
| > You don't have to go to art school to produce art. Even
| if you do go, there's nothing stopping you from producing
| art unlike what you studied.
|
| Sure, but also you do, presumably, have to _see_ to
| paint. The overwhelming majority of our training corpus
| is just visual observation.
| thfuran wrote:
| What are you trying to suggest is the significance of
| that?
| chx wrote:
| Human art can exist without anything prior. For example,
| I highly doubt the creators of cave paintings traveled
| the world to see other such work before doing theirs. But
| you could look at impressionism or pointillism for that
| matter.
|
| AI can not.
|
| Indeed, the problem is the utter unoriginality of the
| current wave of AI. Literally the only thing it can do is
| to produce the next token which is the most likely to be
| _in the training set_.
| FeepingCreature wrote:
| I doubt the creators of cave paintings stayed their
| entire lives in that cave. One presumes they saw animals.
|
| Could an agentic AI generate a cave painting if you
| restricted it to training data of animals and also
| limited its ability to painting with primitive tools?
| Sure, why not?
| dahart wrote:
| What is this 'agentic AI' you speak of? There's no such
| thing to date. There is no AI yet that can train on
| pictures of animals and then _decide_ to produce cave
| paintings. If you limit the output to be produced only
| with primitive tools, then the AI is not agentic, you're
| just making a fixed-function paint machine.
| dahart wrote:
| Maybe because that's misleading and not very useful?
| There are different kinds of stealing, and the platitude
| "all art is stolen" conflates them. The kind you're
| referring to is stealing ideas, and is not illegal. The
| kind parent is referring to is what copyright is designed
| for, and it happens to be illegal. It's helpful to
| distinguish between these two kinds of stealing, because
| one can be good for society and art and business, and the
| other not so much.
|
| The historical context of the idea that "all art is
| stolen" is intentional hyperbole, it's meant to be
| provocative and a little but humorous, and not meant to
| be taken absolutely literally like you are implicitly
| suggesting. It does not mean literally that thieves
| walked out with the paintings, nor does it mean that art
| progressed because people were hand-copying the Mona
| Lisa. It was referring to varying degrees of using other
| people's art as inspiration, borrowing subject matter or
| composition, mimicking techniques, etc.
|
| None of this is what AI does. AI is trained on fixations
| (to borrow the copyright term) and can only produce
| remixes of fixations. It can't do the kind of stealing
| that T.S. Eliot was talking about. It doesn't steal
| ideas, it only steals pixels.
| sumanthvepa wrote:
| "Good artists copy; great artists steal" - Pablo Picasso.
| By this standard AI must be an amazing artist.
| wizzwizz4 wrote:
| Great artists steal, as in, _make things their own_. The
| computer programs can 't extrapolate without producing
| artefact-dominated visual noise: they can only
| interpolate ("copy"). (Besides, you're affirming the
| consequent.)
| seanhunter wrote:
| See my comment above. People say that Picasso said that
| (because Steve Jobs used to quote it as from Picasso) but
| there's no evidence he said that. It seems it was
| actually TS Elliot originally and then that was "stolen"
| by Igor Stravinsky, who was unquestionably a great artist
| but was saying it about Anton von Webern.
|
| https://quoteinvestigator.com/2013/03/06/artists-steal/
| eesmith wrote:
| > imbued with inaccuracies and errors
|
| Which this one certainly was. Tape drives don't look like
| that nor does the arrangement make any sense.
| amelius wrote:
| Maybe you should install an image-generating reader-plugin
| then.
| falcor84 wrote:
| I'd actually love to have one - is there a good one?
| FpUser wrote:
| I Find the same. Even though the AI image has lots of errors
| with the right instructions it usually carries the intentions
| of the master way better than some stock that can be found in
| reasonable time
| wslh wrote:
| It worked for me. Trying to share it here but there is an issue
| with the share API. Will reply to this message or edit it once
| the sharing is working.
| mateus1 wrote:
| What bothers me about this use case is that AI will generate
| the most generic middle-of-the-road illustration that adds no
| insight or value to the prompt. I'd get a better kick if he
| just wrote a description like "I can only imagine them
| debugging this while sitting beside a room-sized mainframe with
| a plethora of wires and bulbs"
| dang wrote:
| " _Please don 't pick the most provocative thing in an article
| or post to complain about in the thread. Find something
| interesting to respond to instead._"
|
| https://news.ycombinator.com/newsguidelines.html
| ayhanfuat wrote:
| Not explicitly stated in the article but the main contributor to
| the speed-up is not the hardware but the software. Solvers have
| improved immensely.
| anonzzzies wrote:
| Off-topic; Sparcstation 5! I have 40 of them and 15 E450s, all
| working. Such lovely machines. Unbreakable.
| macintux wrote:
| I used 5s early in my career, and I can't recall a single
| failure. I very much miss Sun hardware.
| kaashif wrote:
| Why do you have 40 Sparcstation 5s? Very curious.
| anonzzzies wrote:
| I am a hoarder of things that work forever. I have 1000s of
| old computers that all work. I love this tech that I can
| solder... oldest one is from 1963.
| Cockbrand wrote:
| Oh man, err, person, that reminds me that I need to fix my
| SparcStation Voyager, whose 2.5" SCSI drive broke. I do
| have a 2.5" SCSI2SD already, and I also have install media,
| but it always seems like too much effort. Lovely machine,
| though.
|
| Anyhow, in the late 90s I was given an SS10 (?) clone made
| by Axil, which was my main driver for a year or so, running
| RedHat for Sparc.
| zelos wrote:
| I remember a friend showing me photos from someone who shot a
| sparcstation and proved they were _literally_ bulletproof. I
| can't find them now though.
___________________________________________________________________
(page generated 2024-05-25 23:02 UTC)