@@nocodenoblunder6672 But it IS literally just doing that. LLM's give you back the modified versions of the text that they've been trained on. What you're getting back is mostly the "average response" based on the statistical probabilities in the training data. They're extremely good at memorization. They do some limited reasoning, but it's very difficult to figure out whether the model is reasoning or whether it's just using a memorized knowledge in the moment. On top of that, they're virtually unable to say "i don't know", and will just confidently make stuff up. Pair that with people thinking "the machine must know better", and it's a recipe for a disaster.
I saw post where some Software Engineers were talking about how their coding speed tremendously dropped when the tried coding without the AI code editor they had been using for months. Apparently they were struggling to remember simple things (and not just syntax), and how in general they felt a loss of motivation to learn new things the accept the first AI answer thing that TJ was talking about). Imagine paying a company to lobotomize you like this. Maybe I'm just old and grumpy. Maybe I'm just an ex-creative writer, turned programmer who is pissed of at what tech (the last of 3 things I'm passionate about) is turning into.
Completely agree. I think AI just makes you too dependent on itself, or at least if you use it the way most people(and programmers) do. In the long-term, if you use it for everything(or mostly everything) coding related, you might become a worse programmer than when you started and didn't have AI. Since you saw short-term improvements in coding speed, you commit to it and integrate it into your setup as if it was just another VSCode plugin. I also think this principle applies to way more than just programming. And of course, this benefits the companies that provide LLM-related services. Making you more or less forced to use their product is a great business model for them.
but if ai can take some mental burden off your brain then that is good. Do you still do long division or just use a calculator? The whole premise of society is to stand on the shoulders of the people before you, you dont have to know everything, because you cant.
One quote I read in "The Alignment Problem" really stuck with me: "AI/ML is a fantastic tool, provided the future looks like the past." If you your goal is to change the status quo of some sector, tools built with ML are the absolute last thing you should reach for. The context in the book was a jurisdiction that used an ML system to determine whether to detain someone pre-trial, with the goal of removing human bias. In reality, they encoded human bias into the system and AMPLIFIED it, resulting in far more biased outcomes than before, while removing any mechanism for holding anyone accountable for the biases. In certain domains, AI/ML are a form of "bias laundering" whereby you preserve expand human biases, but remove any reliable accountability mechanism.
My biggest issue with AI right now is the push for it. It feels like the business world has an unhealthy obsession with AI, and as a result i suspect that we will push AI forward at the expense of moral and security concerns/issues.
I actually took notes from this video. It's very rewarding to see your thoughts validated by other people, i agree 100% with everything. Bias in LLMs are a huge concern.
Thanks for sharing your thoughts and opinions on some things folks in the industry aren't really talking about when it comes to AI. Loved "small cabal" and "llmnop."
I missed you TJ. We need more of you on YT. That take is so true and a bit scary. I've been coding for 7 years professionally, and around 12 years total, and I see so many new people being fascinated by LLM's without seeing the issues they bring 🙃
My experience with the LLMs is that when they work it's magical, and I'm truly impressed, but I don't think they will replace programmers anytime soon. The perspective nowadays is very often from the programmers, who already know what they want to achieve, and they try to force the LLM to deliver. For me personally it's super annoying, when I know what I want, and I get wrong answers
5 часов назад+1
Loving these incursions into philosophy from you TJ.
Problem solving never goes out of fashion. In the land of the blind proompt engineers (those who have outsourced their faculty of thinking) the guy with the one eye (one who actually understands the code) will be king.
Good points. I saw a post by a math professor that the math assignments in the field of matrix math appeared to be solved by AI. Ironic since matrix math is the math powering AIs. My primary fear about AI, which I also don't see discussed, goes in a somewhat different direction. The really large models, like Chat-GPT are extremely expensive to train, leaving it to a few wealthy corporations to build those. They control through reinforcement training the rhetoric of the text coming out of this. They control how woke the resulting text is. If we play with the idea that in a few years 50% of all text content is written, or rephrased by an AI - the result is that a few large companies control the rhetoric of written language. I see that as a huge democratic problem.
to be honest the AI we see now is saying "look this is magic", brings investors in, gets people thinking. Where I see LLM's in the future is communicating instructions to another machine. They will always be limited to boundaries, prompting works it's best if grammatically correct so any construct that has a boundary of 0 would be ideal for a LLM to generate a near 100% proficiency, you could train them on individuals, train them with certain disciplines. I think the biggest problem with the big LLM companies is that they want to show us that it is "alive", the "magic", they forgot to build upwards instead of they built downwards. straight to the brain instead of starting with legs first. But then it would not be magic
The LLMs of today will not be able to do anything like that in any given amount of time. If they had infinite time for training, they will plateau ( as they pretty much already did). Just remember one thing, nature gives no free lunch. You trained AI on 40 years of human data. Now you are 3 years in the future. How are you getting another 40 years of human generated data? You can’t. You have a time machine? I don’t think so.
Yeah not to comment three times in a row but you’re literally saying exactly what I’ve been expecting to happen right now. These LLMs are going to funnel non engineers into all manner of services and tools that they don’t understand or need. They’re gonna rack up huge cloud infrastructure bills and all manner of mistakes are going to happen outside of coding itself
I use Aider and APIs extensively (all free and unlimited, shout out to Gemini Experiment and Codestral) and I have to say at the current state AI coding is like you guiding a very dump junior developer, you have to extensively review their code and stop them mid way if they start hallucinating on long context codebase (100k+ tokens). But they're a pretty good as a side kick for scaffolding and dirty work. The system prompt and prompting technique make all the different tbh. With new technologies you have to sometimes give them the entire docs or at least the cheatsheet (in case of Raylib, 6 pages cheatsheet is nothing) and they'll do somewhat fine. You still have to do all the heavy lifting tho, but that's what fun about coding using AIs.
TJ died a week later by a car accident. Just joking. Here in Brazil we have a song Cowboy Fora da Lei - Raul Seixas, and he basically say, "I'm not stupid to be a hero, because they can kill me"... I appreciate your work a lot, great admiration, but be wise and safe for your family. Take care!
The irony is what you describe is what alot (most?) agencies do. E.g. on the payrole of a company like Google so will always promote GCP, or use a CMS like Optimizley again because of the kick backs they get from those companies.
Gotta admit, I really didn’t consider this inevitable comercial bias once these tools become the standard as google became in the past. Research should go into developing open source powerful alternatives to keep these balances in check, establishing a community managed system with transparent reports that help everyone understand how these tools get results. This level of power being guided solely by the interest of capital gain looks like a recipe for even more loss of agency over our lives.
Learning how to program is a slow process, so using LLMs to reason about code that either it or a person has written is fine. For personal projects using LLMs is ok too. The attitude from the developers who are all in without a real interest in programming shows they're not really programmers, they're prompters. And like how it fared with Apple, it will generate a lot of money while also increasingly becoming a niche. Your warnings are very valid. There are good people who kinda use LLMs, which imho is fine, they just have to be cautious of who's in the driver's seat.
LLMNOP is such a good abbreviation and mnemonic by following the alphabet. Large Language Model Nascent Oligopoly of Products. I’m sure you named it this way deliberately ;)
One thing I noticed with their new “instruct this model by giving it context “ is that it will be so biased towards the context, such that everything is. ABOUT the context. If you give it the context to be, idk, you are religious, then every answer will be ( let’s say you ask physics questions) something like “as God made them” or “very inline with your religious traditional views”. Try it with language paradigms oop / dod / fp. Even if something is clearly NOT of that language paradigm, it will say it is just to…please you? Another thing ( kind of the same domain with contexts ) is that, whatever the subject is, it will not only do its usual wrong answers, but then also embed them with the context that you prefer things to be put through, and then it’s like talking to a flat earther. I dont think anyone wants to explain to chatgpt or claude why they’re wrong. You don’t pay to tell the AI it is wrong. This is a serious issue because it wastes A LOT OF ENERGY and the messages can be pretty long. F the copyright concerns. The real concern is that this thing, as an industry, is kind of a Las Vegas on steroids ( in terms of energy consumption per value generated to its user ). I don’t hear any government talking about controlling AI usage. It runs as if we have unlimited resources. We don’t!!! We can’t have 8 billion people pay 200$ / month for o1. We technically have the money, if we spread it to everyone, but realistically, we DO NOT HAVE THE RESOURCES to keep this as a maintainable thing. It’s pretty much a lie, because they can’t really deliver and let’s say that they somehow could, they get energy from MAGIC, then you still have the issue that this thing is barely any good for that kind of money spent on it. People are starving around the world, and we actually do have the resources to give them all food, but we don’t. We feed the money to AI. Great job humanity!
Genuinely love this video but I also have a point: let’s pretend the AI was trustworthy and not only trustworthy but omniscient and could ONLY tell the truth. What if it says a truthful statement that goes against what you know to be true? Like if it says, fully correctly, that JavaScript belongs on the server and you’ll go to hell if you don’t listen, what are you supposed to do? How do you square that circle?
I share very similar thoughts as you. I too think AI will be completely disruptive in the industry but fear people will now mistakenly think they don't have to understand the code LLM generates for them. Very similar to the "Copy Paste" dev we have always had but it will be amplified 100x. The easiest way to stand out from the crowd is to be the person who truly understands. You can still use LLMs to help in your understanding but you have to approach it with that mindset. The people who are going to be the most effective at using AI code assistant tools are the ones who still have a fundamental understanding of how the code that gets generated works. I am worried many people are seeing notable figures in tech state how much they love using these tools without realizing that they've had years and years of experience programming before LLMs were a thing. They have developed a level of intuition they can use when evaluating the output of LLMs. You are never going to build this intuition yourself if you blindly tab away.
One thing I always think about is if the LLMs are trained on all publicly available code and if most code is mid wouldn't that mean that LLMs will always be Mid and can't escape that hatch. I mean AI models are very sensitive to their training data(think image generation tools can't generate images of left handed people because the majority of people are right handed) and many such cases in AI. But maybe the new reasoning model paradigm fixes this somehow idk. So if you want a solid future pick a deprecated stack and language so it's not well represented in the dataset 😂 There isn't a better time to learn Fortran
You don't even need regulation to block the way ... The cost of GPUs that can run LLMs is quite high !! e.g: even the upcoming Digits computer for $3k can't run Deepseek. You'll need 7+ to run it. And the limit of stacking them is 2 for now. Any way, $3k is still very high for a student or most people!
Great points! It seems that LLMs become just another no-code tool? Like really, what is the different between a non technical person using something like bubble vs an LLM to generate an application.. the point being here that no code tools have not replaced developers..
AI will do all the jobs, with average performance, average of past results as it was in 2022, when AI started to feed on itself and everything went to shit. You use it for everything that's not important, for important things you become expert and if you're able to be better than the average of all humans, you're going to be better than AI.
1) LLM's only work off of massive amounts of EXISTING training data examples created by real programmers. 2) LLM's take over 3) Real world programmers stop creating new real world examples 4) NOTHING NEW in programming is EVER created again.
it will be like accounting. Accounting, like programming, has a standardized set of rules to follow, it is not exactly difficult to do and there is automated software for it. Small businesses just use automated accounting software to do basic accounting, but once you get into large corporations, there is more creativity needed by a human accountant.
Eh, this isn't quite that (although the same concerns do exist in such a possibility). These are simply risks of adopting AI to automate some programming tasks. Using AI of the current non-AGI variety will accelerate the homogenization of programming solutions for given tasks. It's like an even more powerful instance of the problem we have with frameworks. You wind up with the majority of programmers not really knowing how to do things outside of the scope of the framework. In the right hands (a skilled programmer's) AI will continue to be a game changer, but what we've got now is not building full applications on its own outside of a very limited scope.
The main risk in my opinion is that people trust AI as if it was intelligent.
This. However as well to some degree.
Well its not like its complelty dumb either, which some people make it out to be. "Oh its JUST predicting the next token not a big deal"
as if it _were_
@@nocodenoblunder6672 But it IS literally just doing that.
LLM's give you back the modified versions of the text that they've been trained on. What you're getting back is mostly the "average response" based on the statistical probabilities in the training data. They're extremely good at memorization.
They do some limited reasoning, but it's very difficult to figure out whether the model is reasoning or whether it's just using a memorized knowledge in the moment. On top of that, they're virtually unable to say "i don't know", and will just confidently make stuff up.
Pair that with people thinking "the machine must know better", and it's a recipe for a disaster.
Lol I know. But the Average American IQ is 98. There's just not much we can do about it.
I saw post where some Software Engineers were talking about how their coding speed tremendously dropped when the tried coding without the AI code editor they had been using for months. Apparently they were struggling to remember simple things (and not just syntax), and how in general they felt a loss of motivation to learn new things the accept the first AI answer thing that TJ was talking about).
Imagine paying a company to lobotomize you like this. Maybe I'm just old and grumpy. Maybe I'm just an ex-creative writer, turned programmer who is pissed of at what tech (the last of 3 things I'm passionate about) is turning into.
Completely agree. I think AI just makes you too dependent on itself, or at least if you use it the way most people(and programmers) do.
In the long-term, if you use it for everything(or mostly everything) coding related, you might become a worse programmer than when you started and didn't have AI. Since you saw short-term improvements in coding speed, you commit to it and integrate it into your setup as if it was just another VSCode plugin.
I also think this principle applies to way more than just programming. And of course, this benefits the companies that provide LLM-related services. Making you more or less forced to use their product is a great business model for them.
but if ai can take some mental burden off your brain then that is good. Do you still do long division or just use a calculator? The whole premise of society is to stand on the shoulders of the people before you, you dont have to know everything, because you cant.
One quote I read in "The Alignment Problem" really stuck with me:
"AI/ML is a fantastic tool, provided the future looks like the past."
If you your goal is to change the status quo of some sector, tools built with ML are the absolute last thing you should reach for.
The context in the book was a jurisdiction that used an ML system to determine whether to detain someone pre-trial, with the goal of removing human bias. In reality, they encoded human bias into the system and AMPLIFIED it, resulting in far more biased outcomes than before, while removing any mechanism for holding anyone accountable for the biases.
In certain domains, AI/ML are a form of "bias laundering" whereby you preserve expand human biases, but remove any reliable accountability mechanism.
It all comes down to: “why learn math when there are calculators which do math better than you”.
Life isn’t about convenience but it’s about learning.
Quote: "LLMs don't seek truth." Exactly! Great point! Thank you!
Well, a lot of people don't seek truth as well :) But 💯
My biggest issue with AI right now is the push for it. It feels like the business world has an unhealthy obsession with AI, and as a result i suspect that we will push AI forward at the expense of moral and security concerns/issues.
The prompters are alchemists, the programmers are chemists.
This is a good take. The analogy to a degraded google search with “ads” is interesting.
Good to hear some rational concerns aired... rather than just "ai ate my hamster" ... nice work.
Agree 1000%. A lot of wishful and magical thinking going on amongst our "thinking class".
I actually took notes from this video. It's very rewarding to see your thoughts validated by other people, i agree 100% with everything. Bias in LLMs are a huge concern.
Thanks for sharing your thoughts and opinions on some things folks in the industry aren't really talking about when it comes to AI. Loved "small cabal" and "llmnop."
Hey VS C**e fan here and I share your concerns. This definitely needs to be considered and discussed more. Good take 👍
Also I've watched other videos from you which I liked but I also liked this style of video from you too
I missed you TJ. We need more of you on YT.
That take is so true and a bit scary. I've been coding for 7 years professionally, and around 12 years total, and I see so many new people being fascinated by LLM's without seeing the issues they bring 🙃
My experience with the LLMs is that when they work it's magical, and I'm truly impressed, but I don't think they will replace programmers anytime soon. The perspective nowadays is very often from the programmers, who already know what they want to achieve, and they try to force the LLM to deliver. For me personally it's super annoying, when I know what I want, and I get wrong answers
Loving these incursions into philosophy from you TJ.
Its very unusual to see TJ don't make funny about something that I feel strange watching and thinking: "This is TJ?"
Great points, I think similar!!
what are you using for the slides? I forgot what its called. Can anyone tag it here?
Problem solving never goes out of fashion. In the land of the blind proompt engineers (those who have outsourced their faculty of thinking) the guy with the one eye (one who actually understands the code) will be king.
Teej with the elite takes!!!
Add to that cli-ai. You make an update to your server, and you find out someone is getting busy with your cpu and ram.
I'm glad there's at least some real competition in this space or this would be an even bigger concern.
Good points. I saw a post by a math professor that the math assignments in the field of matrix math appeared to be solved by AI. Ironic since matrix math is the math powering AIs.
My primary fear about AI, which I also don't see discussed, goes in a somewhat different direction. The really large models, like Chat-GPT are extremely expensive to train, leaving it to a few wealthy corporations to build those. They control through reinforcement training the rhetoric of the text coming out of this. They control how woke the resulting text is. If we play with the idea that in a few years 50% of all text content is written, or rephrased by an AI - the result is that a few large companies control the rhetoric of written language.
I see that as a huge democratic problem.
I really like this content, great video!
to be honest the AI we see now is saying "look this is magic", brings investors in, gets people thinking. Where I see LLM's in the future is communicating instructions to another machine. They will always be limited to boundaries, prompting works it's best if grammatically correct so any construct that has a boundary of 0 would be ideal for a LLM to generate a near 100% proficiency, you could train them on individuals, train them with certain disciplines. I think the biggest problem with the big LLM companies is that they want to show us that it is "alive", the "magic", they forgot to build upwards instead of they built downwards. straight to the brain instead of starting with legs first. But then it would not be magic
How did you make this presentation?
This happens in other industries when a few players decide what is the mainstream.
title "Oracle" still gives me headache and bad flashbacks after working with this database
LLM's training humans and not the way around in not so distant future is my greatest fear.
The LLMs of today will not be able to do anything like that in any given amount of time. If they had infinite time for training, they will plateau ( as they pretty much already did). Just remember one thing, nature gives no free lunch. You trained AI on 40 years of human data. Now you are 3 years in the future. How are you getting another 40 years of human generated data? You can’t. You have a time machine? I don’t think so.
"AI will enslave us all!"
"Meh"
"AI will make Typescript the only programming language!"
"Oh no!"
5:35 yes, the same way we trusted the benevolent doctors running the WHO in 2020, right ?
How do you make that slides?
AI burnout goes brrr
I thought teej was gone for another 6 minutes, but he is back
Yeah not to comment three times in a row but you’re literally saying exactly what I’ve been expecting to happen right now. These LLMs are going to funnel non engineers into all manner of services and tools that they don’t understand or need. They’re gonna rack up huge cloud infrastructure bills and all manner of mistakes are going to happen outside of coding itself
There’s gonna be so many people that start businesses with the outputs from LLMs and wind up in huge debt from networking fees and other shit
A RAG module that scores paid placements would be trivial (and could emerge without direct intent by the plagiarizers-in-charge)
so like ads thru rag?
Vercel's v0 also seems to fall into the first and last problems you mentioned in terms of dogfooding/self-promo.
I use Aider and APIs extensively (all free and unlimited, shout out to Gemini Experiment and Codestral) and I have to say at the current state AI coding is like you guiding a very dump junior developer, you have to extensively review their code and stop them mid way if they start hallucinating on long context codebase (100k+ tokens). But they're a pretty good as a side kick for scaffolding and dirty work. The system prompt and prompting technique make all the different tbh. With new technologies you have to sometimes give them the entire docs or at least the cheatsheet (in case of Raylib, 6 pages cheatsheet is nothing) and they'll do somewhat fine. You still have to do all the heavy lifting tho, but that's what fun about coding using AIs.
Bro beeped vscode🤣
glad you appreciated that :)
recovered my brain from we-are-doomed-state into fun-state
Watch your language, man, there's junior engineers here 😂. You can't just mention ****de without censoring.
TJ died a week later by a car accident. Just joking. Here in Brazil we have a song Cowboy Fora da Lei - Raul Seixas, and he basically say, "I'm not stupid to be a hero, because they can kill me"... I appreciate your work a lot, great admiration, but be wise and safe for your family. Take care!
My partner was talking to an llm, and it responded to everything she said! I’m Fckd!
Would be hillarious if with time llms become just bunch of if statement with hardcoded sponsored products
The irony is what you describe is what alot (most?) agencies do. E.g. on the payrole of a company like Google so will always promote GCP, or use a CMS like Optimizley again because of the kick backs they get from those companies.
Introducing: sponsored LLMs knowledge totally not behind users backs.
love the ppt in vim
Which presentation tool did you use for this?
teej! how do I too get my terminal to do an slideshow? :))
Gotta admit, I really didn’t consider this inevitable comercial bias once these tools become the standard as google became in the past.
Research should go into developing open source powerful alternatives to keep these balances in check, establishing a community managed system with transparent reports that help everyone understand how these tools get results.
This level of power being guided solely by the interest of capital gain looks like a recipe for even more loss of agency over our lives.
We wanted LMFAO but we got LLMs
Learning how to program is a slow process, so using LLMs to reason about code that either it or a person has written is fine. For personal projects using LLMs is ok too. The attitude from the developers who are all in without a real interest in programming shows they're not really programmers, they're prompters. And like how it fared with Apple, it will generate a lot of money while also increasingly becoming a niche. Your warnings are very valid. There are good people who kinda use LLMs, which imho is fine, they just have to be cautious of who's in the driver's seat.
It’s all about capture.
Wait, I love typescript
11:06 It's more accurately calledCorporatocracy.
If you've seen "Idiocracy", you know exactly what I mean with Brawndo practically being the govt.
LLMNOP is such a good abbreviation and mnemonic by following the alphabet. Large Language Model Nascent Oligopoly of Products. I’m sure you named it this way deliberately ;)
One thing I noticed with their new “instruct this model by giving it context “ is that it will be so biased towards the context, such that everything is. ABOUT the context. If you give it the context to be, idk, you are religious, then every answer will be ( let’s say you ask physics questions) something like “as God made them” or “very inline with your religious traditional views”. Try it with language paradigms oop / dod / fp. Even if something is clearly NOT of that language paradigm, it will say it is just to…please you? Another thing ( kind of the same domain with contexts ) is that, whatever the subject is, it will not only do its usual wrong answers, but then also embed them with the context that you prefer things to be put through, and then it’s like talking to a flat earther. I dont think anyone wants to explain to chatgpt or claude why they’re wrong. You don’t pay to tell the AI it is wrong. This is a serious issue because it wastes A LOT OF ENERGY and the messages can be pretty long. F the copyright concerns. The real concern is that this thing, as an industry, is kind of a Las Vegas on steroids ( in terms of energy consumption per value generated to its user ). I don’t hear any government talking about controlling AI usage. It runs as if we have unlimited resources. We don’t!!! We can’t have 8 billion people pay 200$ / month for o1. We technically have the money, if we spread it to everyone, but realistically, we DO NOT HAVE THE RESOURCES to keep this as a maintainable thing. It’s pretty much a lie, because they can’t really deliver and let’s say that they somehow could, they get energy from MAGIC, then you still have the issue that this thing is barely any good for that kind of money spent on it. People are starving around the world, and we actually do have the resources to give them all food, but we don’t. We feed the money to AI. Great job humanity!
If AI can create a WordPress App, that would be revolutionary. Except for that, not gonna use it to write code
Genuinely love this video but I also have a point: let’s pretend the AI was trustworthy and not only trustworthy but omniscient and could ONLY tell the truth. What if it says a truthful statement that goes against what you know to be true? Like if it says, fully correctly, that JavaScript belongs on the server and you’ll go to hell if you don’t listen, what are you supposed to do? How do you square that circle?
But you can tell the llm what tools to use.
I share very similar thoughts as you. I too think AI will be completely disruptive in the industry but fear people will now mistakenly think they don't have to understand the code LLM generates for them.
Very similar to the "Copy Paste" dev we have always had but it will be amplified 100x.
The easiest way to stand out from the crowd is to be the person who truly understands. You can still use LLMs to help in your understanding but you have to approach it with that mindset.
The people who are going to be the most effective at using AI code assistant tools are the ones who still have a fundamental understanding of how the code that gets generated works.
I am worried many people are seeing notable figures in tech state how much they love using these tools without realizing that they've had years and years of experience programming before LLMs were a thing. They have developed a level of intuition they can use when evaluating the output of LLMs.
You are never going to build this intuition yourself if you blindly tab away.
One thing I always think about is if the LLMs are trained on all publicly available code and if most code is mid wouldn't that mean that LLMs will always be Mid and can't escape that hatch. I mean AI models are very sensitive to their training data(think image generation tools can't generate images of left handed people because the majority of people are right handed) and many such cases in AI. But maybe the new reasoning model paradigm fixes this somehow idk.
So if you want a solid future pick a deprecated stack and language so it's not well represented in the dataset 😂
There isn't a better time to learn Fortran
A year or two ago I asked chat gpt about luau and it thought it was typescript... might be better now tho but real concern.
You don't even need regulation to block the way ... The cost of GPUs that can run LLMs is quite high !! e.g: even the upcoming Digits computer for $3k can't run Deepseek. You'll need 7+ to run it. And the limit of stacking them is 2 for now.
Any way, $3k is still very high for a student or most people!
preach!
shots been taken on his business partner/ friend
Great points!
It seems that LLMs become just another no-code tool? Like really, what is the different between a non technical person using something like bubble vs an LLM to generate an application.. the point being here that no code tools have not replaced developers..
guys, I think teej may be more intelligent and considered than prime.. just saying
well, thanks for the paranoia :)
AI will do all the jobs, with average performance, average of past results as it was in 2022, when AI started to feed on itself and everything went to shit. You use it for everything that's not important, for important things you become expert and if you're able to be better than the average of all humans, you're going to be better than AI.
more like llmNOPE
1) LLM's only work off of massive amounts of EXISTING training data examples created by real programmers. 2) LLM's take over 3) Real world programmers stop creating new real world examples 4) NOTHING NEW in programming is EVER created again.
wtf is "macapitalism"
🙌
AI bias and propaganda similar to film propaganda in 20s
me agree
LLM makes you a another sheep in the flock
Yeah you’re *soo* intelligent for not using it. Good for you sweetie!~
@@yogxoth1959 sometimes you've gotta lurk in the flock
I like how we went from "AI will never replace programmers" to "these are the risks of replacing programmers with AI".
it will be like accounting. Accounting, like programming, has a standardized set of rules to follow, it is not exactly difficult to do and there is automated software for it. Small businesses just use automated accounting software to do basic accounting, but once you get into large corporations, there is more creativity needed by a human accountant.
Eh, this isn't quite that (although the same concerns do exist in such a possibility). These are simply risks of adopting AI to automate some programming tasks.
Using AI of the current non-AGI variety will accelerate the homogenization of programming solutions for given tasks. It's like an even more powerful instance of the problem we have with frameworks. You wind up with the majority of programmers not really knowing how to do things outside of the scope of the framework.
In the right hands (a skilled programmer's) AI will continue to be a game changer, but what we've got now is not building full applications on its own outside of a very limited scope.
The possibility of lobbying is exactly a capitalism problem
First!
Use a spellcheck dipshid