I always think about how relatively bug-free older games like GTA San Andreas are, since if they ship with some game breaking bug it's pretty much permanent. You could say this about software in general, but with games, especially non-linear ones I'd imagine there'd be more edge cases involved (I've never developed a full game myself, so I can only speculate).
In paper mario 64 (20001), there was a game breaking bug where I got enough star points (or whatever they're called) and got the prompt asking me to level up. But I was already levelled up to the max, and the game wouldn't let me proceed without levelling up. I couldn't roll back to the previous save game because every time I beat the boss I'd get enough points asking me to level up again and I'd be stuck. These days a simple patch would do the trick.
I can't imagine a game like Cyberpunk 2077 coming out in 2004 in the state that it did.
You can do months of testing which obviously costs money AND delays your game. Or you can ship it and patch later. Make your customers be the beta testers.
Now I personally wait at least 6 months because I can't fucking STAND broken games.
However the market has spoken: gamers at large don't give a shit.
Unless you are so enraptured by the idea of the game that you cannot live without it, or if you are a streamer who will make money from playing the game on day 1, waiting 6 months for patches is the play.
You will typically save money on the game, other people will review it, flaws will be fixed, and you'll have a better time of it.
One regular workflow is a reconciliation we do for events that we put on — a number of costs that are expensed, a number of costs that are prepaid until the event happens, individual registration revenue that is recognized immediately and then the corporate sponsorships that are often paid in advance but their recognition is deferred until the event happens. Previously since that involved both balance sheet, income statement and CRM reporting, we relied on an integration vendor to write custom scripts to bring all the info (poorly) into our ERP. Since then, we’ve found a tool leveraging LLMs to ‘join’ those various sources and our events people generally described the report they wanted with a template in excel and it readily created that report with “export to sheets or excel” functionality.
A report that previously took ~4 hours per month for a very expensive resource now takes 30s to validate and can be run completely ad hoc by the events managers.
Every evil propaganda corp has a cutesy brand name like that. Names like Voice of the People, or Electoral Freedom Foundation, or Moms for Liberty, or Free America Foundation or First Amendment Recovery Temple (I definitely made that one up, and some of the others).
I don't think those two things are mutually exclusive. Good chance that a few students that cheated or at the least used AI in a major capacity to graduate, still booed when that former Google CEO brought up AI at the graduation speech. Being pro AI when it benefits them and anti AI when it doesn't is just human nature. I'm being a little reductive here though.
Auto-complete on steroids, is still my favorite analogy for AI. I don't mean that in a negative way either. Autocomplete is very good, but that never stopped me from learning English grammar and spelling.
Quite right. I'm worried about the impact that LLMs will have on the learning process, especially in programming, but also in writing. Programming and writing are both skills that seem simple, but take an absolutely staggering amount of practice to master.
Think about how much your own writing (and programming, if you were lucky enough to start early) evolved from, say, age 12 (when a lot of smart kids start to tackle 'real' books) to age 18 (when you supposedly have a good enough education for 50% of work in most countries) to age 25.
All of that evolution is a direct result of one thing: practice! But with a magic answer box available in everyone's pocket, it'll take truly Herculean effort from a learner to actually grind through the practice instead of just cheating for an answer. I really worry how much an LLM user will actually comprehend their own code or even prose; if you've scarcely written a line of code, how can you really understand what's going on in a debugger? If you haven't done the legwork of writing essays and constructing coherent arguments and comprehending grammar, how will you ever communicate effectively?
Maybe I'm just a dinosaur and these kids will sail a whole level of abstraction above my own understanding of writing and programming, much like how my own generation preferred Python to C, and how the previous generation evolved from assembly to C/BASIC/etc. But then I come back to those missing fundamentals, that empty mental model. It's not like my English or CS teachers had me grind through essays and implementing linked lists and Djikstra's Algorithm for pure busywork. They did it because practice is the only way to truly immerse a student in a practical subject. Maybe it'll work for programming, as long as LLMs get good enough that you can always ask them to fix low-level errors for you? But it seems unlikely to work in prose. And even those generational programming jumps I mentioned (assembly to C to Python) were lossy; most kids I went to school with would be absolutely useless writing C code, and even as a bit of a dinosaur I'm pretty awful at even debugging assembly.
Like you said: you still need to learn grammar and spelling. And I suspect a whole skill tree of other fundamentals!
One angle I'm exploring, as a non-dev who nonetheless works in tech, is using Claude as a professor. Make learning timelines for me for Leetcode, break it down in phases, start with theory, ask me questions, then give me a coding challenge. Save that to an html artifact I can export and read on my phone.
It still gets things wrong, I can tell as I get through problems.
But it was either that or that dreary 'Cracking the Coding Interview' book. At least I'm learning fundamentals by asking question after question and making it track the concepts I had trouble with.
That's one use. Will most people use it to learn? Probably not. But most people are ... most people.
Yup, I used to believe that people would all use the Internet to educate themselves, and we all know how that turned out (loads of people did, but the majority didn't).
The way I think of it has evolved a lot over the last 5 years. At this point I think human brains probably do something analogous to next token prediction when we think. For all the hype, I think LLMs are actually more, not less, intelligent than that average person realizes. I think it’s legit, actual intelligence, not just “artificial” intelligence. That may be a hot take but it’s just my perception.
> At this point I think human brains probably do something analogous to next token prediction when we think
That's reasonable, but it doesn't mean that LLMs are close to being brains.
For a start, when humans think/talk, we often think ABOUT something - whatever is swirling about in our mind, or what we are currently seeing/feeling/etc. An LLM generating tokens/words is doing so only based on it's weights and the word sequence it is currently generating ... the human parallel would be more like a rapper spitting out words based on prior words, essentially on auto-pilot, or when we get triggered into spitting out stock phrases like "have a nice day".
If you want to compare an LLM to a human brain, it's basically equivalent to our language cortex if you ripped out all the external connections and ripped out all the feedback paths that make it capable of learning.
Of course there is a lot more to our brain than just our language cortex, but that alone should make you realize there is no real comparison beyond the fact that our language generation is also going to be based on prediction, and partly auto-regressive.
Having shame would require the LLMs to actually be able to recognize mistakes they make.
People love to put a lot of meaning on what an LLM responds with when asked why it made a mistake, but it's critical to remember that the answer to that prompt is just another series of probabilistic tokens, and has no actual relation to how the error happened.
They "recognize" mistakes just fine because you explicitly tell them. They recognize them well enough to correct (...sometimes). The way in which mistakes don't register is "Oh shit, that bad result was a result of my inappropriate actions. I must pay attention to not doing that again or the user will think I'm an idiot. I should even think about it some more to avoid the whole class of mistakes". Think of emotions as an attention mechanism that LLMs lack.
And many times, after an error is pointed out and an LLM offers an "explanation" for what happened, the LLM then gives the exact same erroneous result.
It's language. Language itself is the thing that makes us smart in the unique way that we are among the other animals, and it weirdly turns out to be transferable to machines to at least some degree.
At least 50% of humans have no "inner voice" and are not thinking in the same way as you. Many animals like dolphins, dogs, rats, crows are also very intelligent yet appear to only have primitive language capabilities.
A lot of human intelligence is really societal rather than individual, based on knowledge transmitted down through generations by writing (the real enabler). If you take that away then what you are left with is something more like an isolated hunter-gather tribe.
I personally think that the "inner voice" is a non-falsifiable claim, and therefore more of a religious belief than something which can be part of any materialist theory. In this regard, I'm a strict empiricist and wouldn't be able to claim that I have one myself. In fact, I find that thinking "out loud" or "on paper" produces much better results in most instances, probably because I'm grounding my thinking in natural language, which is a fantastic medium for thought. If my "inner voice" were comparable in efficacy to actually speaking or writing, we wouldn't notice this effect, but I'm definitely not alone in this regard.
Your point about writing and social intelligence is, to me, more evidence for the "it's language that's smart, not us" hypothesis. We start off in small bands of hunter-gatherers that store their intelligence in an oral culture. Language then jumps to clay tablets, papyrus, codex books, etc. The printing press allows it to escape containment to a wider public than just a caste of priests and bureaucrats. As soon as we invent automatic calculators, we start networking them and using those to process language, albeit in a primitive way (email, the web, etc.). Recently we discovered some abstruse math that, with the assistance of a bunch of beefy video cards, can crunch centuries of human writing into a mathematical object that encodes at least some of the meaning of that writing into an even more "advanced" symbolic processing machine. There's a clear trajectory of language itself getting more and more free of the specific wetware it grew up on.
It's a falsifiable claim, in that if there is a way to train a useful LLM from scratch without any human authored input language to bootstrap it (something I've been on the lookout for but haven't seen, though admittedly I'm not an AI researcher, just some Linux nerd with a day job as an SRE), then we can disprove it.
For the religious angle, look no further than John 1:
"In the beginning was the Word, and the Word was with God, and the Word was God."
Well, humans developed language. Language is just a tool that let's us leverage our innate intelligence.
I'm sure that we will eventually build artificial brains, capable of bootstrapping communications and language for themelves (if run en-masse in a simulation where the benefit of communication would emerge). An LLM can't do this since it is by definition/construction something only capable of learning a pre-existing language.
An artificial brain, just like a wet jiggly one, is always going to be more intelligent than a one-trick pony like an LLM - a language processor, but it is notable how intelligent that one-trick pony nonetheless appears to be.
I think it's interesting that you think we could bootstrap an artificial brain with no inputs from human culture. I disagree, but am open to an existence proof of this kind. Such an artificial brain would be totally alien to us, of course. I wonder how differently it would perform versus something more grounded in "real" culture and writing?
I'd say that human babies and LLMs are both existence proof that prediction and prediction error feedback is all you need to learn. The artificial brain/baby would be designed to learn just like us by prediction, and should therefore be capable of learning language from scratch just as we do.
You could choose to lock it in a virtual or physical basement with printout of the Common Crawl dataset and raise it like an LLM that learns language with zero real world grounding, in which case it may feel a bit like an LLM (but smarter - able to learn, etc), or you could let it interact with the real world and learn everything, including language, that way and be a lot more grounded.
It's hard to guess how the grounded version would feel to talk to vs the CommonCrawl one - I think it would mostly come down to how far we wanted to go in making the artificial brain have all the moving parts of a human one. In an odd way the more human we tried to make it, the more alien it might feel, but not for the "uncanny valley" reason you might imagine...
The thing with an LLM that makes it feel so human is that they are designed to 100% copy humans - their output is 100% driven by the training goal of trying to exactly match the samples in the training set. As soon as we start to try to build something more brain-like then it's behavior is going to be a lot less predictable - not just auto-regressive "auto-pilot speech", but driven by it's own internal thoughts, emotions, innate traits, etc - depending on how much of our brain we tried to copy in the artificial one. I expect it would feel a lot more organic, less robotic, to talk to, but at the same time perhaps less human since unlike the LLM it's not built just to mimic human speech.
Yes, this to me is also a good sign for the "it's language that's smart, not us" argument. It's an emergent trait that has evolved several times, like flight or carcinization. There's something about language that attracts evolution toward it. One would expect such a trait to have a big survival value (disclaimer: IANA biologist, philosopher, theologian, mathematician, or linguist).
> I think it’s legit, actual intelligence, not just “artificial” intelligence. That may be a hot take but it’s just my perception.
You might be redefining words here; there isn't a form of intelligence that isn't actual intelligence. It is all actual intelligence. Artificial in this context means it is something we're creating in a lab. LLMs can't avoid being artificial intelligence. The meaning of "AI" is to artificially create actual intelligence.
average person is absolutely awful judge on anything you put in front of average person tho.
And if anything, average AI user is vastly overstating how good/useful it is. Papers about it pretty much always show huge gap between "productivity person thinks they are achieving" and "actual growth of productivity"
I guess an optimistic way to look at this would be to treat this as just another layer of abstraction, meaning people could focus on larger scale problems moving forward, similar to how the evolution of programming languages influenced development time, quality and the quantity of software being put out. The question is at what price does all of this abstraction come at, assuming AI continues to evolve at its current rate.
This can not be seen as layer of abstraction as it's non deterministic and not trustworthy. So we still need to inspect and understand that abstraction layer output if we want to have a reliable product
I wonder how Steve Jobs would've reacted to this GenAI boom. He constantly talked about the intersection of Humanities and tech, as well as fostering creativity by pushing people to their limits (for the better or worse), so I don't think he'd be one of those CEOs that's first in line to get rid of human workers as much as possible. Or maybe he would be and I'm just giving him too much credit.
On an unrelated note, I haven't used an Iphone since 2018 and I wonder if Siri has gotten any better. I do see "Apple Intelligence" being advertised everywhere and besides AI summaries of texts on the notifications bar I haven't seen anything to understand what Apple Intelligence actually means.
It's just a broad term for whatever AI integration they put into their various Apps and services. So, a combination of the neural engine stuff they've been doing for years, and integration with white label AI services from Google or OpenAI.
Siri is basically unchanged, it looks like they have had serious problems getting LLMs, or generative AI in general to be reliable and 'safe' enough to put their own name on it. By 'safe' I mean thinks like not generating emails based on Mein Kampf, or doodles of genitals, or hallucinating false 'facts'.
Not a concern for many of the frontier AI providers with no reputation to burn, but not exactly on-brand for Apple. I very much doubt Jobs would have viewed that differently.
Hallucinations and other errors are a real problem. I’ve tried tinkering with Shortcuts (which get you more direct access the models) to generate AI responses to messages and emails, but it’s unreliable. The biggest problem is it has trouble distinguishing between people, confusing me with the person I’m conversing with, or with third parties.
How good is AI integration in Apple products? Did they drop the ball as hard as Microsoft did? I naively assumed a few years ago that Microsoft could pull it off perfectly because they had more than enough in terms of resources & engineers (yes, I was this naive in college)
Like most MS vs. Apple differences, it comes down to a matter of taste. They've added quite a few AI enhancements across their apps and operating system, but they are mostly feature enhancements and not major AI branded efforts. Having a Summarize button in Mail.app where it's contextually relevant or having text improvement menu options in text fields vs. slapping a major "Copilot" tab into everything.
Their use of AI so far has been much less "let AI take the wheel and brand it as a product itself" and more "use AI to improve an aspect of <user need>".
It’s very limited. Only a few buttons to push, unless you roll up your sleeves. For instance there’s a button to have it summarise an email, but having it search through all your emails requires writing scripts.
Yeah, hard to guess how a person would react to transformative technology, together with whatever context it'd be brought up, their reaction could differ.
I too would say Jobs probably would have an human angle on it, but he also famously was a tyrant who struggled with people not doing exactly what he asked, and could be slightly nitpicky about that, maybe having a robot that follows exactly what he wrote, to a fault, would be a machine he'd greatly enjoy.
Or he'd throw it in the trash with some flourish of words explaining how a machine could never feel frustrated so therefore couldn't great excellent products, or something.
His reaction probably still would not have been solidified yet, given how long his response took to other tectonic shifts in technology. That isn’t to say he wouldn’t have an opinion to voice, I just suspect it wouldn’t have resulted in a product direction for at least a few more years.
> I wonder how Steve Jobs would've reacted to this GenAI boom.
Steve Jobs really cared about his users, and putting out great products for those users.
I imagine he would have loved all the machine learning stuff that Apple has being doing the past few years (stuff like voice noise separation, instant text OCR and photo object isolation).
Based on the story about the first iPod being too big, dropping a prototype in a fish tank, lots of air bubbling up and him going "there's your space", or the disdain he displayed about how crappy Mobile.me was, I imagine he would have recognized LLMs for the flakey product they are and would have been very wary of introducing them into users their workflow.
> .. and I wonder if Siri has gotten any better ..
Siri is still crap, but so is Gemini. Both still do incredibly stupid stuff like when you try to request some music on Spotify "cannot find the artist or song 'My Playlist Hard Techno'" / play some unknown vaguely matching artist. Or it'll do an internet search for "goose oven cooking timer ten minutes". Or ask "for how long should I set your timer?" and name the timer "goose oven cooking timer ten minutes" which in a way is even more stupid.
You'll get some naysayers here saying stuff works perfectly, but its that inconsistency that sucks. Sometimes it'll one-shot a really difficult voice command or obscure song search. And then other times (many times..) I have to yell at it three times to set a timer, at which point I sigh, realize doing it manually would've been faster, and set the timer manually.
In a way its made me realize LLMs and voice assistants aren't that good, it's just that even tech people have incredibly low standards. Especially the people working in AI.
The problem is natural language is a horrendously bad human-computer interface. Even if they're running nondeterministic software, computers are very precise machines. You wouldn't talk to your lathe or milling machine and expect good things to happen. So why would you have that expectation of a computer? It's ridiculous sci-fi fantasy nonsense.
It's hilarious, when you boil away all the froth and hype, that we've collectively decided that "talk to computer" is somehow worth an entire generation of venture capital and maybe even the whole stock market. It's a dumb idea to begin with. A mouse and keyboard are better.
Which, to their credit, seems to be what Apple tried to do with Apple Intelligence and was already doing with Machine Learning. But if under Steve they had over promised and under delivered—like what happened under Cook—some heads would probably have rolled.
> I wonder if Siri has gotten any better.
Nope. There are rumours the new one will use Gemini and be better, but who knows. We’ve heard this before.
> I haven't seen anything to understand what Apple Intelligence actually means.
When it was announced, I thought it was a brilliant piece marketing in the sense of associating the “A” in AI with Apple. But then it turned out to be trash, so turns out the association is a hindrance. Anyway, you know how Microsoft uses “Copilot” for anything they ship which has “AI” in it? That’s Apple Intelligence. It’s the umbrella term for anything anywhere in one of their products where they use any kind of AI/ML.
The difference is it’s incredibly easy to opt out of apples AI-like services. For instance, I have never had Siri on on my iPhone no matter how many years go by. And every time I’ve gotten a new one, it stayed off. One tap, that’s it.
They don’t go out of their way to bolt the features to everything the phone does or make it particularly difficult to turn them off. That’s probably one of the last major reasons I still have an iPhone.
Microsoft in comparison forces you to use OneDrive, has copilot tapping on your glass like clippy every five seconds, etc. The desperate pleas to use these features are embarrassing
Good point. Yes, I also have all of that turned off and can ignore it. Not that Apple is without its pushiness, though, I still get frequent nags to “upgrade” to Tahoe and iOS 26 without an option to turn those off, and every time I go to System Settings to update to a new macOS 15 version and click to show more details, they sneakily select macOS 26 checkbox.
no this is a fair question. he was enough of a sociopath to disown his own kid, but his narcissistic tendencies and love of the arts would have been a weird counter point to that.
