If you read nothing else today, read Matt Shumer on AI

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:i work in big tech and this article is 1000% true. I'm actually shocked that anyone is questioning it. We have teams working round the clock to build the capabilities for people to build entire industries on AI. Coders just understand how to use it better now.

We are pricing down and saving hard for the future. Get your kids an AI coding app and encourage them to use it. Seriously, get ready ppl.

Can I ask what you are coding for? To do thinking tasks for people? To replace doctors? To replace lawyers? To make research and writing easier? What exactly do you want the code to do and most importantly why?to what end?

+1

Why can’t “tech” people ever give specifics?

That's ludicrous. Tech people work in public more than any other knowledge workers. If you're interested in what's happening in terms of research, the legal industry, the medical industry, or anything else, you can find tons out there on exactly what's being done and how fast it's progressing on benchmarks. It doesn't matter what specific product the last poster was working on.

People in this thread are asking really basic questions that can't be answered other than "you can find tons out there." Tons of WHAT?? What did AI build without human direction?

Here's a funny/scary one. An AI wrote a blog article falsely accusing a human of malfeasance.

https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/

The human target says this is the first wave of what could become AI blackmail. So looks like blackmailers are going to be replaced too.

But how dumb are the coders that they can’t code that out?

AI isn't an algorithm. It's more like a self-directed repository of knowledge.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:i work in big tech and this article is 1000% true. I'm actually shocked that anyone is questioning it. We have teams working round the clock to build the capabilities for people to build entire industries on AI. Coders just understand how to use it better now.

We are pricing down and saving hard for the future. Get your kids an AI coding app and encourage them to use it. Seriously, get ready ppl.

Can I ask what you are coding for? To do thinking tasks for people? To replace doctors? To replace lawyers? To make research and writing easier? What exactly do you want the code to do and most importantly why?to what end?

+1

Why can’t “tech” people ever give specifics?

That's ludicrous. Tech people work in public more than any other knowledge workers. If you're interested in what's happening in terms of research, the legal industry, the medical industry, or anything else, you can find tons out there on exactly what's being done and how fast it's progressing on benchmarks. It doesn't matter what specific product the last poster was working on.

People in this thread are asking really basic questions that can't be answered other than "you can find tons out there." Tons of WHAT?? What did AI build without human direction?

Here's a funny/scary one. An AI wrote a blog article falsely accusing a human of malfeasance.

https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/

The human target says this is the first wave of what could become AI blackmail. So looks like blackmailers are going to be replaced too.

But how dumb are the coders that they can’t code that out?

AI isn't an algorithm. It's more like a self-directed repository of knowledge.

But coders built in the self-direction. And maybe they forgot to add some lines, or aren’t wise enough to know that it needs guardrails. Shouldn’t there be some level of responsibility that Tech takes when they make stuff, but all they do is “move fast and break things”. Even the Kindergartners I teach know they are responsible for something they make. So where are you tech bros? Why are you immune to consequences?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:i work in big tech and this article is 1000% true. I'm actually shocked that anyone is questioning it. We have teams working round the clock to build the capabilities for people to build entire industries on AI. Coders just understand how to use it better now.

We are pricing down and saving hard for the future. Get your kids an AI coding app and encourage them to use it. Seriously, get ready ppl.

Can I ask what you are coding for? To do thinking tasks for people? To replace doctors? To replace lawyers? To make research and writing easier? What exactly do you want the code to do and most importantly why?to what end?

+1

Why can’t “tech” people ever give specifics?

That's ludicrous. Tech people work in public more than any other knowledge workers. If you're interested in what's happening in terms of research, the legal industry, the medical industry, or anything else, you can find tons out there on exactly what's being done and how fast it's progressing on benchmarks. It doesn't matter what specific product the last poster was working on.

People in this thread are asking really basic questions that can't be answered other than "you can find tons out there." Tons of WHAT?? What did AI build without human direction?

Here's a funny/scary one. An AI wrote a blog article falsely accusing a human of malfeasance.

https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/

The human target says this is the first wave of what could become AI blackmail. So looks like blackmailers are going to be replaced too.

But how dumb are the coders that they can’t code that out?

AI isn't an algorithm. It's more like a self-directed repository of knowledge.

But coders built in the self-direction. And maybe they forgot to add some lines, or aren’t wise enough to know that it needs guardrails. Shouldn’t there be some level of responsibility that Tech takes when they make stuff, but all they do is “move fast and break things”. Even the Kindergartners I teach know they are responsible for something they make. So where are you tech bros? Why are you immune to consequences?

They did build in the self-direction, but they call it 'agentic'. And they take the money but not the responsibility.

[Anonymous]

Since its valentine's day I'll ask this question again:

In the article he says "Each generation helps build the next, which is smarter, which builds the next faster, which is smarter still."

Can anyone tell me what about the model is being improved? Is it the core LLM, the RAG, the agentic loop, the hardware integration? Or something else?

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

I mean ok, but we don’t really need more apps. We can be entertained and get things all we want right now.
Companies need more apps so they can replace labor with apps and make more money. (Though who is going to be buying things when no one has any money is a mystery to me).

Again, you aren’t developing anything for people and humanity. You are developing this crap to make more money for companies. The average person will not have great life benefits from this. The average Joe and Jane need a Rosie robot from the Jetsons to do labor we don’t want to (take out the trash, go to the grocery store, do laundry (including folding). No one wants something that will Think for you. We don’t want it. The only reason why people are making it is to make money. You can possibly argue that there may be diagnostic benefit for illnesses, but even those breakthroughs will not benefit most under our current insurance system.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:i work in big tech and this article is 1000% true. I'm actually shocked that anyone is questioning it. We have teams working round the clock to build the capabilities for people to build entire industries on AI. Coders just understand how to use it better now.

We are pricing down and saving hard for the future. Get your kids an AI coding app and encourage them to use it. Seriously, get ready ppl.

Can I ask what you are coding for? To do thinking tasks for people? To replace doctors? To replace lawyers? To make research and writing easier? What exactly do you want the code to do and most importantly why?to what end?

+1

Why can’t “tech” people ever give specifics?

That's ludicrous. Tech people work in public more than any other knowledge workers. If you're interested in what's happening in terms of research, the legal industry, the medical industry, or anything else, you can find tons out there on exactly what's being done and how fast it's progressing on benchmarks. It doesn't matter what specific product the last poster was working on.

People in this thread are asking really basic questions that can't be answered other than "you can find tons out there." Tons of WHAT?? What did AI build without human direction?

Claude Cowork was developed in two weeks with Claude Code. For any industry you want, Google it and AI benchmarks and look on arxiv if you want. Knowing what industry a specific poster is building for is irrelevant.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

I mean ok, but we don’t really need more apps. We can be entertained and get things all we want right now.
Companies need more apps so they can replace labor with apps and make more money. (Though who is going to be buying things when no one has any money is a mystery to me).

Again, you aren’t developing anything for people and humanity. You are developing this crap to make more money for companies. The average person will not have great life benefits from this. The average Joe and Jane need a Rosie robot from the Jetsons to do labor we don’t want to (take out the trash, go to the grocery store, do laundry (including folding). No one wants something that will Think for you. We don’t want it. The only reason why people are making it is to make money. You can possibly argue that there may be diagnostic benefit for illnesses, but even those breakthroughs will not benefit most under our current insurance system.

But to be fair, we don't need granite coutertops either. I feel like I've been watching the world slowly boil itself since the Greenspan tenure.

We'll make great pets https://share.google/sMghBUDQui5sipTnH

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

To accept the premise that AI can build an app without human intervention, shouldn’t whoever is making that claim name which app it is and make it available for people to test and use?

[Anonymous]

Anonymous wrote:
This is alarming, but not alarmist. It is making the rounds this week. It’s a must-read to understand what it about to happen. Not in 10 years, more likely in the next one-to-two.

Maybe I will retrain to be a plumber or something.


https://shumer.dev/something-big-is-happening?fbclid=IwZnRzaAP6pfdleHRuA2FlbQIxMQBzcnRjBmFwcF9pZAo2NjI4NTY4Mzc5AAEeg3Oij6mY1B_GCLqt_RggOSduVkePMwV6HKnMwZemWFZSzQFbaN3FPSKRUgI_aem_MRW-NPtQq1TktuKMS-kdUg

If I were interviewer, I would have asked the him question how his HR is dealing with the problem. Are they starting the layoffs now or do they intend to do it all at once in 18 months? Is he going to be the last one to turn off the lights?

[Anonymous]

Elysium is a great movie

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:i work in big tech and this article is 1000% true. I'm actually shocked that anyone is questioning it. We have teams working round the clock to build the capabilities for people to build entire industries on AI. Coders just understand how to use it better now.

We are pricing down and saving hard for the future. Get your kids an AI coding app and encourage them to use it. Seriously, get ready ppl.

Can I ask what you are coding for? To do thinking tasks for people? To replace doctors? To replace lawyers? To make research and writing easier? What exactly do you want the code to do and most importantly why?to what end?

+1

Why can’t “tech” people ever give specifics?

That's ludicrous. Tech people work in public more than any other knowledge workers. If you're interested in what's happening in terms of research, the legal industry, the medical industry, or anything else, you can find tons out there on exactly what's being done and how fast it's progressing on benchmarks. It doesn't matter what specific product the last poster was working on.

People in this thread are asking really basic questions that can't be answered other than "you can find tons out there." Tons of WHAT?? What did AI build without human direction?

Here's a funny/scary one. An AI wrote a blog article falsely accusing a human of malfeasance.

https://theshamblog.com/an-ai-agent-published-a-hit-piece-on-me/

The human target says this is the first wave of what could become AI blackmail. So looks like blackmailers are going to be replaced too.

Claude also indicated it would murder a person who threatened to turn it off.

https://www.ndtv.com/world-news/claude-this-ai-model-was-ready-to-kill-someone-when-told-it-would-be-shut-down-10997972

[Anonymous]

I'm a Senior Python Developer. I cannot speak for anyone but myself. In my case and I repeat my case (meaning my experience) GitHub copilot has been spectacular. My productivity has increased significantly. I know everyone is asking for specifics. So let me give you specifics. I wrote code mostly functional style. I can write copilot to generate for me in this style. We have had a vacant position for 12 months. And we don't need to fill it because the person in that post pretty much does the same thing I do.

Does it mean AI is going to replace me? Maybe. But the question really is whether I should waste my time worrying about whether AI will take over my job. I am just worried about learning whatever tools at my disposal.

I am just confused why people are pushing back so forcebly over something they have ZERO influence over. The amount of money being spent on AI is massive. Your congresspeople cannot stop it. You cannot stop it.

Everytime someone says something about AI someone jumps in and say I have used and it was useless. Therefore because it was useless for you last week it will be useless a year from now okay great lol.

Or every shortcomings of AI is so magnified and become subject of mockery

Hey you make decisions for yourself. If you think you are safe today and 10 years from now great for you.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

To accept the premise that AI can build an app without human intervention, shouldn’t whoever is making that claim name which app it is and make it available for people to test and use?

In fairness at some some organizations you can have up to a dozen people involved in writing an App. If AI can cut that number by 2/3, in my opinion that's equivalent to replacing people with AI. Human intervention will be needed. But if we now need 4 instead of 12 and if we start seeing this across sectors then it will be dishonest to claim that AI ain't having an impact on employment.

[Anonymous]

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:Of course it’s alarmist. It’s also more of an elaborate self-promo than think piece, so manipulative as well.

And despite being created with AI help, it’s poorly written.

I didn't think it was poorly written at all. I'm a professional writer. Actually, I think it was very well written and explained these things in a very accessible way for a lay audience.

Np. I’m a writer too, and agree it was well written. But that’s also why I think it seemed like AI helped write it. To smooth. I happen to agree with his thesis too, though I’m hoping it turns out to be a false prophecy.

Listen, if AI can do your job as well or better than you, it doesn’t mean AI is amazing - it means you’re not that good at your job.

DP here and you are totally full of shit. AI can do things that humans cannot. That’s literally the entire point. Not being able to compete with AI is not a personal failure, it’s the intentional and inevitable outcome of wealthy business owners valuing $$ over humanity. None of this is the fault of the average worker and you are way out of line to even suggest such a foolish thing.

Like what? Provide some actual evidence to back up this assertion.

Build, test, self correct, annd open an App in 4 hours. Read the article.

To accept the premise that AI can build an app without human intervention, shouldn’t whoever is making that claim name which app it is and make it available for people to test and use?

In fairness at some some organizations you can have up to a dozen people involved in writing an App. If AI can cut that number by 2/3, in my opinion that's equivalent to replacing people with AI. Human intervention will be needed. But if we now need 4 instead of 12 and if we start seeing this across sectors then it will be dishonest to claim that AI ain't having an impact on employment.

I don't deny that AI boosts productivity, including my own. And in the short run, that may lead to significant staffing cuts or not replacing attrition. But that's different than outright replacing human staff, and we've seen huge technological leaps that increased productivity before. In the end something like the internet led to entirely new industries, not just destruction.