> By that logic, any technology that you can get certified in is too complicated?
In IT, I am inclined to agree with that. In real engineering, it's sometimes necessary, especially dangerous technology and technology that people trust with their life
> dangerous technology and technology that people trust with their life
Software runs on so many things we depend on IMO it also in many cases falls in the "dangerous technology" category.
Non-hobby OSes, non-hobby web browsers, device drivers, software that runs critical infrastructure, software that runs on network equipment, software that handles personal data, --IMHO it would not be unreasonable to require formal qualifications for developers working on any of those.
Claude, come up with a protocol for communicating between AI agents and IDEs/editors. Create node, python and rust libraries. Create a website with a landing page
Honestly, I'm tempted to see if Gemini can write a Sublime Text plugin that implements this protocol.
Feels like a lot of mindshare has shifted towards VSCode, so that's where the tooling investment goes. I don't want to be forced off of subl because new tools stop supporting it - it's a great editor, and it's not sponsored by a massive company.
As you suggest, I've had a moderately successful time trying to get AI to write its own Sublime Text plugins so our favorite editor doesn't get left behind, so might be cool to try with this too?
Shouldn’t be too hard. I wrote this emacs plugin to do similar things but without this protocol - https://github.com/kgthegreat/protagentic . Heavily used AI assisted coding for it.
An init system without the ability to specify dependencies? Without user/group configuration? Ordering must be manually configured? No parallel service launching? No resource management?
Please don't call this an init systern. It's a barebones process supervisor.
It actually does all these things. Quite well, even - in my experience better than systemd.
I didn’t use nitro, I’ve been using daemontools (which nitro is an evolution of) for decades. Incredibly easy to use, incredibly stable, understand, and control.
There is no well defined way to do dependencies (what if your dependency dies 3 seconds into the process? There are many right answers). The djb/daemontools way is just “it’s your problem. But here are the reliable simple cheap tools to start, stop and monitor your dependencies”.
It is lightweight, very easy to understand and reason about (much more than systemd - and it’s not just an issue of familiarity - it has much fewer moving parts than systemd’s service management part).
It’s rock solid. It goes to great length to never lose process output, not even a single char, across service restarts. (it might be possible to achieve same on systemd - but it isn’t trivial)
And it’s been that way for me for two decades now - Ubuntu moved from system v to upstart to systemd; my systems still use the same daemontools setup they used 15 years ago, they do it on FreeBSD and Linux. And they just work, no surprises.
Had to buy an IPv4 address for a VPS the other day in order to clone some git repositories. Couldn't believe it. Costing their customers money when they should be able to support v6 by now.
They charge €0.50 per month to add an IPv4 address. A shared IPv4 NAT gateway introduces a whole lot of problems for them just to support customers who need IPv4 but don't want to pay a tiny amount for it.
How would a server-side NAT know which Hetzner customer it should route a request to? It has an encrypted packet arriving at this shared address on port 443. You can route a shared address to the proper service based on the HTTP Host header but that can only be done by the customer using their encryption key, so no sharing an address between customers. Home LAN NAT only works because the router can change the source port used by the request so that responses are unambiguously routed to the right client.
I don't think they're saying they should support incoming connections on such a NAT, I think they're saying that servers behind the NAT would be able to make outgoing connections (e.g. to access shared resources).
In regards to an EC2, AFAIK, not necessarily. You pay extra for an elastic IP (IPv4) which is the equivalent to a static IP but the EC2 is assigned an IPv4 address and an IPv6 when IPv6 is enabled.
I'll show you a few misspelled words and you tell me (without using any tools or thinking it through) which bits in the utf8 encoded bytes are incorrect. If you're wrong, I'll conclude you are not intelligent.
LLMs don't see letters, they see tokens. This is a foundational attribute of LLMs. When you point out that the LLM does not know the number of R's in the word "Strawberry", you are not exposing the LLM as some kind of sham, you're just admitting to being a fool.
Damn, if only something called a "language model" could model language accurately, let alone live up to its creators' claims that it possesses near-human intelligence. But yeah we can call getting some basic facts a "feature not a bug" if you want
So people that can't read or write have no language? If you don't know an alphabet and its rules, you won't know how many letters are in words. Does that make you unable to model language accurately?
So first of, people who _can't_ read or write have a certain disability (blindness or developmental, etc). That's not a reasonable comparison for LLMs/AI (especially since text is the main modality of an LLM).
I'm assuming you meant to ask about people who haven't _learned_ to read or write, but would otherwise be capable.
Is your argument then, that a person who hasn't learned to read or write is able to model language as accurately as one who did?
Wouldn't you say that someone who has read a whole ton of books would maybe be a bit better at language modelling?
Also, perhaps most importantly: GPT (and pretty much any LLM I've talked to) does know the alphabet and its rules. It knows. Ask it to recite the alphabet. Ask it about any kind of grammatical or lexical rules. It knows all of it. It can also chop up a word from tokens into letters to spell it correctly, it knows those rules too. Now ask it about Chinese and Japanese characters, ask it any of the rules related to those alphabets and languages. It knows all the rules.
This to me shows the problem is that it's mainly incapable of reasoning and putting things together logically, not so much that it's trained on something that doesn't _quite_ look like letters as we know them. Sure it might be slightly harder to do, but it's not actually hard, especially not compared to the other things we expect LLMs to be good at. But especially especially not compared to the other things we expect people to be good at if they are considered "language experts".
If (smart/dedicated) humans can easily learn the Chinese, Japanese, Latin and Russian alphabets, then why can't LLMs learn how tokens relate to the Latin alphabet?
Remember that tokens were specifically designed to be easier and more regular to parse (encode/decode) than the encodings used in human languages ...
But actually, you can see an intense enough source of (monochromatic) near-UV light, our lenses only filter out the majority of it.
And if you did, your brain would hallucinate it as purplish-blueish white. Because that's the closest color to those inputs based on your what your neural network (brain) was trained on. It's encountering something uncommon, so it guesses and present it as fact.
From this, we can determine either that you (and indeed all humans) are not actually intelligent, or alternatively, intelligence and cognition are complicated and you can't conclude its absence from the first time someone behaves in a way you're not trained to expect from your experience of intelligence.
Being confused as to how LLMs see tokens is just a factual error.
I think the more concerning error GP makes is how he makes deductions on fundamental nature of the intelligence of LLMs by looking at "bugs" in current iterations of LLMs. It's like looking at a child struggling to learn how to spell, and making broad claims like "look at the mistakes this child made, humans will never attain any __real__ intelligence!"
So yeah at this point I'm often pessimistic whether humans have "real" intelligence or not. Pretty sure LLMs can spot the logical mistakes in his claims easily.
Your explanation perfectly captures another big differences between human / mammal intelligence and LLM intelligence: A child can make mistakes and (few shot) learn. A LLM can’t.
And even a child struggling with spelling won’t make a mistake like the one I have described. It will spell things wrong and not even catch the spelling mistake. But it won’t pretend and insist there is a mistake where there isn’t (okay, maybe it will, but only to troll you).
Maybe talking about “real” intelligence was not precise enough and it’s better to talk about “mammal like intelligence.”
I guess there is a chance LLMs can be trained to a level where all the questions where there is a correct answer for (basically everything that can be benchmarked) will be answered correctly. Would this be incredibly useful and make a lot of jobs obsolete? Yes. Still a very different form of intelligence.
> A child can make mistakes and (few shot) learn. A LLM can’t.
Considering that we literally call the process of giving an llm several attempts at a problem "few-shot reasoning", I do not understand your reasoning here.
And LLM absolutely can "gain acquire knowledge of or skill in (something)" of things within its context window (i.e. learning). And then you can bake those understandings in by making a LoRa, or further training.
If this is really your distinction that makes intelligence, the only difference between llms and human brains is that human brains have a built-in mechanism to convert short-term memory to long-term, and llms haven't fully evolved that.
If I had learned to read utf8 bytes instead of Latin alphabet, this would be trivial. In fact give me a (paid) week to study utf8 for reading and I am sure I could do it. (yes I already know how utf8 works)
And the token/strawberry thing is a non-excuse. They just can't count. I can count the number of syllables in a word, regardless of how it's spelled, that's also not based on letters. Or if you want a sub-letter equivalent, I could also count the number of serifs, dots or curves in a word.
It's really not so much that the strawberry thing is a "gotcha", or easily explained by "they see tokens instead", because the same reasoning errors happen all the time in LLMs also in places where "it's because of tokens" can't possibly be the explanation. It's just that the strawberry thing is one of the easiest ways to show it just can't reason reliably.
> When you point out that the LLM does not know the number of R's in the word "Strawberry", you are not exposing the LLM as some kind of sham, you're just admitting to being a fool.
I'm sorry but that's not reasonable. Yes, I understand what you mean on an architectural level, but if a product is being deployed to the masses you are the fool if you expect every user to have a deep architectural understanding of it.
If it's being sold as "this model is a PhD-level expert on every topic in your pocket", then the underlying technical architecture and its specific foibles are irrelevant. What matters is the claims about what it's capable of doing and its actual performance.
Would it matter if GPT-5 couldn't count the number of r's in a specific word if the marketing claims being made around it were more grounded? Probably not. But that's not what's happening.
I think we’re saying the same thing using different words. What LLMs do and what human brains do are very different things. Therefore human / biological intelligence is a different thing than LLM intelligence.
He's the boss of the researchers so he knows more than them /s
But seriously tho, what parent is saying isn't a deep insight, it makes sense from a business perspective to consolidate your products into one so you don't confuse users
I use Claude Code authenticated via the API (Anthropic Console). There's no limits for me. And I also assume API-metered requests are prioritized, so it's faster as well.
The API does have limits but they’re determined by your monthly spend. I did a trial of tier 1 spend and did hit the limits, but on on tier two spending it was much much better.
In IT, I am inclined to agree with that. In real engineering, it's sometimes necessary, especially dangerous technology and technology that people trust with their life
reply