Socrates' objections to writing wasn't that it was inherently bad, but that it introduced limitations; namely, that you couldn't have a discussion with the author of a text while reading it, and therefore, reading was inferior to talking.
It could be argued that AI is the first step in 2000+ years towards addressing this specific problem.
This could eventually completely transform our understanding of Antiquity. It is estimated that only around 1% of the ancient works in Greek and Latin have survived to the present day, much less in other languages such as Punic [0]. Some works and some authors we only know by name because they were alluded to in later texts.
It's also well known that surviving texts survived because they were copied again and again on costly animal skin during the Middle Ages, by monks who had to make a choice and naturally favored topics that were of most interest to them.
I'm not sure what you expect to find that would completely transform our understanding of the time period. The most likely discoveries are going to be filling in details about things we already knew. This period of time was already pretty well documented. Even if we found something amazing like some of Aristotle's lost works, we basically already know what they were and (roughly) what was in them. Really the most interesting and useful finds would be more mundane things like household records, and personal diaries.
Disagree. My understanding is that most surviving works have been transcribed repeatedly over the centuries, often times based on preferences of the people living at the time. There’s a big chance that excavation could find deeply heterodox stuff, I think.
I think it's less that the stuff would be considered heterodox, as just not as good/relevant. Like certain texts were used in the Roman world for school, just kind of universally taught to the literate class. The Aeneid was one of these but before it was written the Annales by Ennius was the classic poem everyone had to learn. Then the Annales became less popular, stopped being taught, and now we only have some fragments of it.
> There’s a big chance that excavation could find deeply heterodox stuff, I think.
Heterodoxy (or really, orthodoxy) wasn't really a thing in 79ad, and you're not likely to find much of it in the private library of a wealthy Roman's vacation home. The only forbidden work you're going to see from that era is stuff critical of the emperor.
Epicureanism was fairly heterodox/countercultural even during its heyday, and our sources on it are much more limited than the sources on more acceptable schools such as Stoicism. For example, we don't really have any writings from Epicurus except for short fragments, when we know from his students that he wrote many books. Much more survives from his students, but even then one of the main sources of our knowledge of Epicureanism (especially before we started recovering Herculaneum scrolls) was Seneca, a Stoic writing about it as a rival school. None of this was forbidden at the time, but it was unpopular (especially among the ruling class) and eccentric (ditto).
It's hard to imagine something more heterodox than Ovid (he managed to get himself exiled by Augustus), and that survived. Medieval readers didn't seem to mind that sort of thing in Greco Roman writing, it was part of their heritage, no one was seriously worshipping those gods so it wasn't seen as a threat. The people in the past behaving in a way that was seen as immoral wasn't a problem.
Could change our understanding of history - slavery, early Christianity, politics, secularism among roman elite, etc
Or of technology- steam power, mechanical computation (like the Antikythera mechanism, which is the only known example of such a thing until 1300 years later), mechanized production, mining techniques, etc
For our understanding of Christianity, the Pompeii eruption was too soon to the events, so unless the owner was amongst the first Christians worldwide and there is a dated letter from St Luke foretelling the destruction of the temple I don't think it'll be too revolutionary
We've seen this happen already once with the recovery of palimpsests. Outside of a few lucky discoveries, the vast majority of what monks were discarding were things that were not seen as useful - outdated (to them) legal texts, liturgical books, etc.
The exception though would be Greek literature. Greek literacy collapsed in the early medieval era and a large catalogue was probably just scrapped or discarded before even being collected in Monasteries. Herculaneum could represent a legitimate treasure trove in that regard.
Precisely how early Heliocentrism began to be seriously considered is a very much open question, but of its more lucid proponents, very little survives, if anything at all. Usually only in snippets told by others.
And that's just one thing, who knows what else those old Greeks/Phoenicians/etc were kicking about.
You can disable language checks under the paragraph or character tab for the text item (I forget which). A little annoying but at least makes them go away.
> Every memory begins with tiny changes inside the brain
Maybe. But the brain is not the only place where memory is stored. Flat worms remember things (and skills!) after their head has been cut off and they regrew it:
"Every memory begins with tiny changes inside the brain. A discovery that helped explain those changes has earned neuroscientist Oswald Steward one of science’s highest honors.
Steward received the 2026 Kavli Prize in Neuroscience, a USD 1 million award and one of science’s most prestigious awards, for research that transformed scientists’ understanding of how the brain learns and stores memories."
And that's what it took. One comment on hackernews and the prize was retracted. HN at its best! ;)
Well, it does seem that memories may be embedded in the nervous system as well as the brain, so I don’t think the OP is wrong. You sometimes hear of heart transplant patients having other people’s memories / preferences. So, it’s not good evidence, but it’s a possibility.
> Personality changes have been reported following organ transplantation. Most commonly, such changes have been described among heart transplant recipients. [...] A cross-sectional study was conducted in which 47 participants (23 heart recipients and 24 other organ recipients) completed an online survey. In this study, 89% of all transplant recipients reported personality changes after undergoing transplant surgery, which was similar for heart and other organ recipients.
Who knows the cause though, could be anything I suppose, not necessarily that "memory sits in tissue".
It's massive, hugely traumatic surgery, taking the patient past what was considered the point of death a century ago, bringing them back alive, and all with the aid of some of the most powerful drugs in modern medicine's arsenal.
And if your heart is needing transplantation in the first place, you'll be running far below optimal for blood O2 and a dozen other things.
It'd be more surprising if it didn't result in significant change.
> if your heart is needing transplantation in the first place, you'll be running far below optimal for blood O2 and a dozen other things.
that's not obvious on the face of it from a mechanical standpoint. I'd you're sick enough to need one, wouldn't we just give you a mechanical temporary replacement, and keep that one going as high as possible instead as low as possible?
AFAIK mechanical hearts can be prone to creating blood clots and can put more stress on other parts of the body, so it's not necessarily the safest option to just run it flat out.
Just 47 participants in the online study and a wide variety of psychological issues: “depression, anxiety, psychosis, and sexual dysfunction.”. A few of them found religion.
Is anyone surprised that surviving traumatic surgery with a long debilitating recovery time causes mental stress?
> Is anyone surprised that surviving traumatic surgery with a long debilitating recovery time causes mental stress?
No, but some of the references reports and studies (referenced in the above study), if truthful, would be too much of a coincidence, no?
> a heart and lung transplant at Yale-New Haven hospital in 1988. Following surgery, Sylvia developed a new taste for green peppers and chicken nuggets, foods she previously disliked. As soon as she was released from the hospital, she promptly headed to a Kentucky Fried Chicken to order chicken nuggets. She later met her donor’s family and inquired about his affinity for green peppers. Their response was, “Are you kidding? He loved them… But what he really loved was chicken nuggets”
> a 5-year-old boy received the heart of a 3-year-old boy but was not informed about his donor’s age or cause of death. Despite this lack of information, he provided a vivid description of his donor after the surgery: “He’s just a little kid. He’s a little brother like about half my age. He got hurt bad when he fell down. He likes Power Rangers a lot I think, just like I used to. I don’t like them anymore though” (p. 70, [8]). Subsequently it was reported that his donor had died after falling from an apartment window while trying to reach a Power Ranger toy that had fallen onto the window ledge. After receiving his new heart, the recipient refused to touch or play with Power Rangers
I'm sure there might be other explanations to all of these, but at least people are trying to study it more.
Drugs can alter taste. I've had experience with cancer patients, and a side effect of the treament is that food starts to taste differently. Would not be surprised if the drug+treatment combo of traumatic surgery might result in something similar in some percentage of the patients.
Having a suddenly functioning heart probably does wonders for hormone levels and consequently personality.
I once realized that the “snowball” of dread/fear in my chest was likely a minor irregular heartbeat (that I’ve always had) exacerbated in stressful situations.
what do we know about memory other we think in our brains so that's where consciousness must be stored? What do we know about memory that makes that necessarily a leap? What is so special about semantic memory?
Flatworms branched off our side of the animal tree of life very early on. They're on the same side as molluscs, some of whom (cephalopods) are famous for having a more distributed nervous system.
Granted though many/most organs are stateful and somewhat adaptive - in a sense they'll "remember" what happened. Even plants possess that to varying degrees.
That's interesting, but bear in mind that sensory neurons are basically just transducers that sense something and convert it into a neural output. It makes sense that a woman's brain would find it useful to know when she is ovulating.
AFAIK the gut in large terrestrial vertebrates has its own nervous system that rivals the complexity of the entire system in simpler creatures.
The idea that all stateful/regulatory stuff is entirely localised to the brain is a bit too simple to be true. Most of it, sure, but that last few percent can be doing all sorts of clinically important stuff. Nature is an incredibly brilliant engineer, but not always a tidy one.
“A memory begins with tiny changes inside the brain” as truth statement is a basic fallacy of naive physicalism. There is no falsifiable way to ascertain in which direction causality points, nor is natural science even intended provide a definitive answer—it is designed to make predictions and any models that arise in the process are necessarily faulty and do not describe the true nature of underlying reality, which this ultimately comes down to.
Another flatworms-are-fun story: If you irradiate a flatworm using a nuclear reactor, to kill off its stem-cell equivalents, it sort of "rots" as other cells eventually die and aren't being replaced. If after zap, you insert donor stems, its life can instead proceed normally, and the entire body is eventually refreshed. So ship of Theseus like, you have an individual flatworm, swimming through its life, which ends up being genetically unrelated to its earlier self.
People love to bring this up, but while yes, all cells have some form of memory mechanism or another, with quite the number of nerves outside the brain, we are still a cognitively top-eown organism and quite unlike simpler organisms where the line is blurry.
In machine learning terms, we've basically got a layer or two of neurons responsible for physical memory stored outside the brain, sure. But it's out of dozens if not hundreds total.
I don't like talking to strangers and I would consider myself rather introvert, although not extremely (I'm more of a misanthropist, maybe). That said, talking to strangers is really quite easy; I do it sometimes, esp. to entertain my friends while walking on a busy street. It's quite fun. It can happen that you plunge into deep conversations too, with someone you met just seconds ago!
He has a lot of wild defense arguments; one of my favorites is: at some point in his life he lost the ability to speak; to recover his voice he trained it by reading aloud some books over and over, so much so that the content of these books became part of his own brain / of himself.
(Another one, unrelated, but also wild, argues that people who attack him are in fact against science itself, that they want to go back to the Middle Ages, etc.)
It's very obvious he pieced together interesting ideas from others to pass them as his own. And it worked very well, he has radio shows and TV shows and whatnot. And he still has a lot of supporters!
> It's very obvious he pieced together interesting ideas from others to pass them as his own. And it worked very well, he has radio shows and TV shows and whatnot. And he still has a lot of supporters!
Also, being able to get ideas, synthesise them and present them in a way palatable for mainstream audiences is a useful skill and an important role in society. It’s just not research.
> Plagiarizing from people on your own thesis committee is a wild move.
Fun fact: he's using this to prove he didn't do anything wrong, as in "see? the people on my thesis committee didn't care I copied their own work, why should anyone else?"
The truth is, people on "thesis committee" don't read thesis. Some do. The director usually does, if he has the time. But many don't; they glance at the intro and conclusion and call it a day.
> He wrote the thesis at a time when it was impossible to identify lightly rephrased statements across a wide body of works. Now we can dump all of these documents into an LLM and have similar sentences surfaced for human review very quickly
He also uses this to say it's unfair to punish him now with tools that didn't exist when he did the crime, which I find quite rich. If you murdered someone before DNA testing was available, that doesn't exonerate you in any way.
A lot of "plagiarism" is not plagiarism. Feed stuff you wrote into those tools and it will call you a plagiarist every day because you wrote something similar to the person you learned it from.
I don't know about this case, but a lot of these kinds of cases truly are witch-hunts. It's not at all like the reproducibility crisis and faked data and images.
The very few cases that result in sanctions are generally horrendously flagrant.
With another professor I caught a flagrant case in a student thesis and we faced attacks from the university administration because the student had a stellar transcript (also not the positive signal some might think). Punishment was almost inexistent.
It's difficult for me to imagine what it would take to get a doctoral thesis revoked.
You need more than that. No university is going to revoke anything without very good reasons, they have too much to lose. Their first action is always to try to bury the case.
If some in your experience erred on the side of leniency, then it stands to reason that others might err just as egregiously in the opposite direction.
In fact, your anecdote suggests erring is the norm. We should thus expect punishments to be inappropriate in one direction or another. An appropriate punishment seems rather unlikely.
>It's difficult for me to imagine what it would take to get a doctoral thesis revoked.
No respect for the plagiarist physicist, but an easy way to control what media representatives of scientific disciplines get to say publically, is to start out with what amounts to "academic compromat" (scientific fraud, plagiarism, ...).
Did this physicist / media star recently say something controversial?
I mean why did the system let him pass as a physicist, and why did it let him rise the media rank?
> Did this physicist / media star recently say something controversial?
Not really. This is the consequence of an investigation by some journalists about a decade ago, and an audit that lasted for almost 2 years.
> I mean why did the system let him pass as a physicist, and why did it let him rise the media rank?
He is a smooth talker and by all accounts good at vulgarisation. He does well in interviews and is easy to deal with for journalists. There’s always been controversies but media thrive on those.
This goes well beyond accidentally triggering a plagiarism detector.
> Feed stuff you wrote into those tools and it will call you a plagiarist every day because you wrote something similar to the person you learned it from.
The examples in the article use very distinctive wording. One or two occurrences would be forgivable as coincidence or inspiration. An entire document full of examples points to something else.
It seems like that should be the case yet when I listen to any same group of people over a period of time, I often find that those unfamiliar with a concept or solution on day 1 end up repeating it as if it was their own a few weeks later. When I was younger I tended to assume there was an element of intentional theft, but I'm not sure it's natural and a prerequisite to educational acquisition that people can categorize original origin of ideas that may have bounced around them for a long time before they understood their significance.
Sure a common answer would be intentionally copying in the same sessions, less likely is intentionally copying via eidetic memory.. But how much of a spectrum could there be in the middle for memory that would result in repeating a "plagiarism" form months later, etc?
People say how obvious the parlor trick is when they look at a small model LLMs. Well, I've seen the same parlor trick in students who get good grades but seem weak at thought from fundamentals. It seems quite possible to me that in some examples we are now going after them because the environment changed. At much earlier points we did actually value the people who could recite even if somewhat brokenly because we lacked random order recital tools.
You’re wrong; academia has never accepted plagiarism of this magnitude. Enforcement is never perfect, but a doctorate is not an undergrad repeating verbally, it is not thoughtless writing. It’s a doctorate thesis for crying out loud, it has to be novel!
Crediting the origin of the idea is the whole point of citing sources. Learning something from someone doesn't mean the idea is yours now. It means that when you repeat that idea, you should cite the original source of the idea.
This is just how scholarship works. It's not needed in the kind of day to day most of us do, but when you're writing a thesis for a PhD, this stuff matters. You're making the argument that you're expanding the totality of human knowledge with your dissertation, and that requires strict source citing to separate your original scholarship from the sources that influenced it.
What are these tools? I often write about stuff on my blog and I know a lot of what I’m writing or thinking about are ideas someone else has come up with (and that I’ve read but not remembered or not read and come up with a poor version of) but bog standard LLM DeepResearch never picks up the things I want.
I imagine any tool that’s good at plagiarism detection would also kill it at this kind of literature research.
An example of something where it worked like this is that I had some ideas around how tribes evolve and so on and wrote them as I could think of them and ChatGPT was able to find that Darwin’s Cathedral had a far better synthesis of various much more rigorous takes on the subject.
> I often write about stuff on my blog and I know a lot of what I’m writing or thinking about are ideas someone else has come up with
These tools compare words, not idea. They would not detect someone copying concepts but coming up with their own words. I guess some specially fine-tuned LLMs could work but I am not aware of a company actually licensing those for plagiarism detection.
> I don't know about this case, but a lot of these kinds of cases truly are witch-hunts.
There have been a lot of plagiarism accusation in his books already. In this case there was an audit and the conclusions are clear. Whole paragraphs copied and pasted word for word without attribution, about a third of the document overall. If anything, this should have happened about 15 years ago.
Having seen plagiarism first hand, sometimes it exceedingly blatant. Like copying from a PDF that was produced via LaTeX — since LaTeX hyphenates words to split them across lines, if you end up keep-ing the hyphenation in, the te-xt reads like this.
I've seen way worse: a Word document submission that preserved the style and fonts of the sources the plagiarer stole from. As in, font "Calibri 14" only appeared in paragraphs nicked from a source entirely written in that font - and the adjoining paragraphs weren't even size 14!!!
Sadly, this idiot won an award before I was able to see their work, so they had the confusion of receiving an award, and THEN being told they were being spanked for unacceptable behavior. Since they were too stupid to hide the most blatant clues, they had a hard time comprehending this duality.
Academia is very broken if even your thesis committee is A) not interested in reading your thesis and B) can't even be bothered to when it is ostensibly their job.
What exactly is the point of dedicating years of your life to create something exactly nobody is going to read?
Most PhD have a few papers before finishing the dissertation. Many times the dissertation is made of a few paper by the author glued together. The papers usually chain, so it's instead of
introduction1 -> main1 -> conclussion1
introduction2 -> main2 -> conclussion2
introduction3 -> main3 -> conclussion3
the thesis is something like
long introduction -> easy example -> main1 -> main2 -> main3 -> main of preprint -> long conclussion
Thesis by publication is only one way, and not even the most common in many fields. I can't access the actual text of this thesis, but the abstract sounds more like a monograph and I don't see any author publications before the thesis that would lead me to think otherwise.
What gus_massa was describing doesn't sound like thesis by publication. TBP usually uses the papers verbatim with some added material to make them cohesive to the overall thesis argument. What they described is more like repurposing existing material in a thesis format, which is indeed very common.
Haven written a Master's dissertation, I think your typo of conclussion for conclusion (suggesting a head injury is needed to slog through it) is perfect.
For both me (physics) and my wife (history), in the American system, both at strong universities, most of our committee members read most of of our dissertations. For her, in a field where thesis by publication is not standard (your thesis is typically revised into your first book), her committee at the defense focused on questions and comments based on the committee's reading of the thesis more than on the actual defense presentation, which is apparently also normal in the field. In part, I expect that's because the thesis is expected to be built into something important post-PhD, and comments are seen as helpful in that process.
For me, it wasn't quite so apparent at the defense, and I don't know that all members read the final thesis carefully, but most of them had already seen me publish or present most of the research previously, often multiple times. I also know that some (and not just my advisor) did read the final thesis very closely. My thesis was only partially thesis by publication, however, which may have influenced this; it does now have a fair number of citations in its own right, which is somewhat unusual for the theses in the field, and potentially seen as awkward (it means there's significant work in the thesis that I never published elsewhere).
As a caveat, the American system (before current crises) does feel like it can have a two-tier system of PhD students who are expected to remain in academia (we both were) and ones who are not, even at strong universities. Expectations, and attention given, can vary considerably. The American system also tends to have larger and more closely involved committees than, for example, the UK/Irish system.
However, for the form of plagiarism discussed here: if someone had sentences from papers I published years ago interspersed in their work, and they weren't particularly notable sentences, I'm not confident I would notice. Depending on citations and what the sentences were, I'm not even sure I'd mind much, for example, if they were essentially copying a model definition.
It's a long time that the incentive and job structure make universities a very toxic environment. Professors are basically running a 40 years race (about from bachelor or master graduation to retirement). It is still amazing that some good comes out of it.
It's very broken, and I'm not sure if it's possible to write everything original given that you're expected to repeat 2/3rds of past research to fill pages when you write your thesis. For a master thesis that was at least 100 pages. For a PhD nowadays each one of those is published as a book. At least it was like that in my engineering department.
It's a philosophy thesis, and unlike STEM or soft sciences (history, linguistics...), they are very light on fact which make them very dry. The will read the introduction, conclusion (which can have more words than a physic thesis), the main thesis that interest them the most, and count on their collegue to read the other main thesis.
Also, very dry, so it's easy to loose focus, and you can read a rephrasing of your own thesis as a "he has the same ideas" (also, if you do that, please reference the author?)
I find a few of the example damning (hje should totally have added a citation and build his argument around it). Most less so, and i understand that a reader could not catch them.
It varies a lot by field, but in many (not all) scientific fields, a PhD thesis is largely a formality these days. Your publication record is what counts. The days where you could get a tenure track faculty position just on the strength of a PhD thesis are long gone.
Depends on the subfields. CS is by publication, number theory varies ("my students can find a stapler" to the dissertation has revolutionary result not published elsewhere)
CS can (but not frequently) have the revolutionary result you mention as well. A candidate Fully Homomorphic Encryption scheme was first detailed in Craig Gentry's thesis, for example. That being said, this is much less common than a
1. literal stapler thesis, or
2. cleaned up version of a stapler thesis (e.g. rewrite of several previous publications to give broader context etc)
that's broadly true, but there are some areas of CS that are at least as close (say at a minimum PL theory). it's also less math heavy than e.g. complexity theory (though that's admittedly smaller). It can also be less math-heavy than learning theory. This all depends on the type of cryptography as well, especially since it can depend on the region (american cryptography is more theoretical than european cryptography, for example).
that's how i understand it. it's a portfolio with front matter, back matter, the papers that got published with some connective tissue between them and maybe some discussion of the things that didn't work out and why.
Early work in any trade is mostly junk, and academia no exception.
But the process of creating that work, engaged throughought that process with those purported to be more practiced, is usually pretty good at seeding enough expertise and confidence that you might be able to proceed more independently and with real novelty, or might at least be prepared to share the trade with others new to it.
That's the point of those years, and so it's more than a little ironic that AI is being used to undermine a practicing expert while simultaneously eroding the traditional process for becoming one by making it so easy to just generate slop and engage with hallucinations than to actually practice writing deep work or engaging with primary sources.
You probably have plenty of novel ideas in early career, but you almost certainly lack the experience and the basic understanding of your field to develop them properly. Most people have exhausted their own ideas by mid-career. But that that point, they should have the skills and the experience to work on the ideas they come across.
(Looking back at my PhD, it's quite amusing how little did I understand. On the other hand, many of the choices I intuitively made turned out to have some value. But in some cases, understanding that properly took a decade of work by other people.)
Your PhD work is an apprenticeship, after which you are expected to work as a journeyman. The masterpiece that qualifies you for independent work as a tenured professor is often called habilitation. Many academic cultures don't have those, because the expectations are so situational that they don't want to formalize them.
That’s how people outside academia see PhDs. Inside academia, everyone has a PhD and it doesn’t really mean very much. It can take decades to really become an expert in a field, and a PhD program usually lasts around 5 years (in the US).
I believe the person was saying that in academia, literally everyone has a PhD, by definition since it's a requirement for the job, so the simple act of having it means nothing in the context of all of the other people that have it. It of course means a great deal since it's what let's you in to the room in the first place. Imagine interviewing 50 people, every single one of whom have an internship on their resume. What they did during their internship matters of course, but the simple act of having had one doesn't differentiate (matter).
I find it rich how fast you are to jump to destroying the entirety of academia in one stroke. It's quite easy to say things we don't understand should not exist, of course I'm guilty of this myself from time to time. Have you done education beyond the bachelor's degree? It's a very different world.
Who does or doesn't have a PhD isn't terribly important in the scheme of things. Inside academia, the job market is highly competitive, and no-one is getting a job just on the strength of a cookie-cutter PhD thesis. Outside academia, it mostly makes no difference to anything whether you have a PhD or not.
If we apply your criteria, I'm not sure if any universities would be left.
I think they mean that a phd doesn't mean much relatively speaking, since everyone around has one so it's less impressive and you're less of an expert when everyone around you is knowledgeable in the same domain.
That's how it was maybe 100 years ago. Now PhD is just another bit of school work. Sometimes people manage to do really great PhD work, but most of the time it's pretty mediocre or straight garbage.
In some ways, people doing research now have it way more difficult than people of the past. They have hundreds of years worth of research to study before they are on top of things and making an original contribution that stands out among the huge amount of research that already exists is really hard. If we want to keep PhD as a proof of meaningful work, then we ought to lengthen the graduate studies considerably. How about a 10 year PhD program, at the end of which you can really say you have mastered the field?
The value of a PhD thesis is the personal intellectual growth you get from putting it together. The end product isn't really the point.
There's a lot to be said about publishing in academia being broken and how nearly all the value comes from 10% of publications, while the rest are garbage spewed out for reasons orthogonal to the advancement knowledge. However, IMHO, none of that really applies to PhD theses.
> The value of a PhD thesis is the personal intellectual growth you get from putting it together. The end product isn't really the point.
This definitely varies by field. For example, there are some branches of linguistics where the big, important new monographs that move the field forward are often PhD theses (though typically the defended diss manuscript will get some very light rewriting and polishing before it appears from a publisher). After that, a scholar's publications over his/her career might be less ambitious and focused more on minutiae.
AI doesn't work like the rest of the tech industry. The cost of selling another license for a software program is approximately zero.
In the case of AI the marginal cost of the next token is not zero, and is in fact probably not going down much with volume, if at all.
So I'm not sure one can argue that scale will solve everything. It's very much like the old adage "we lose money on every sale, but make it up in volume".
It's wild to think how efficient Internet services were prior to AI. The most expensive thing would probably have been something like encoding video. Now you've a substantial portion of a rack dedicated to a user in the case of something like fable
Best analogue we have is probably video streaming. Or maybe more so live streaming. Unless subscription based and limited time events it seems those don't do well. Twitch has lost money for how long? And most smaller players seem propped up in other ways.
So if there is real cost involved things start to look lot worse and might not be overcome. OpenAI is unlikely to be exception for me.
But there is no indication they are losing money on tokens when R&D and other expenses are factored out? The margins on API are likely very high so the higher the volume the more likely they will be able to cover the other mostly fixed costs.
Also, what are they calling "R&D" exactly? If it is training new models, which needs to be done almost constantly and means spending billions on energy and newer GPUs, then it's not really R&D, but rather operating costs.
So? It's also more effort to work everyday to earn a living than simply stealing what you need from your neighbors at gunpoint. But the law's the law.
As a European I'm conflicted because I think this particular set of privacy laws are overreaching bordering on stupid; but "exemptions" for one of the richest corporations on earth would be beyond absurd and infinitely worse.
Why would you use UUIDs a primary keys? Let SQLite use rowids internally (which is automatic and invisible), and have a different (indexed) column with UUID if you need that for publishing the ID somewhere.
It could be argued that AI is the first step in 2000+ years towards addressing this specific problem.
reply