Getting really sick of people scoffing at the privilege of being disrupted.

“But I need to pay rent and eat!”

Never thinking of imperial needs before themselves!

I 100% agree. GPT-4 just deepens the already huge gap between the richest and poorest and will raise another tech giant.

>if you’re anxious about GPT4 its probably because you’re a shitty programmer

surely this is meant partly tongue-in-cheek?

Whenever I hear stories about someone losing jobs due to a new technology, I would say capitalism and economics rather than technology is to blame here. Yes, you lost your job, but it's not like GPT4 made the decision to fire you, it's your boss who found a better alternative and decided that there is no need for he/she to pay you.

I laughed so hard at the "10x times more productive -> do 10x more work". One simply do not have that much work to do. I used to hire people to paint my wall in a day, now with robotics I can let the same amount of people paint my wall 10 times in a day. No. How about I fire 90% of the workers and let the last one paint the wall with the same time.

Make no mistake, GPT-4 is just a beginning. Far more powerful models will come, without any compromise, without any limitations. They will REPLACE your jobs, take away your income, and you will have no choice but starve after you running out of your savings inevitably. It's not about to earn more or less; It's about winner takes it all. Also more specialized models trained at every profession will come.

Midjourney already completely knocked a huge amount of digital designers and workers out of their jobs. I'm talking about a great profit wipe which is currently happening at a lightning speed across the entire digital art industry. One successful artist often needs to spend 10+ years learning to make better art. Now it's 100% irrelevant. Every company I know is switching to Midjourney at god speed, since you do not want to be left behind. Human resources have already become a burden. You don't even want to believe how fast the entire field is shifting to this without turning back and left countless human creators with now worthless skills in dust.

But remember - the current state of Midjourney is just its early stage. I fully expect its output quality completely decimate every human artist on the earth after 1-2 years, maybe 18 months, maybe even less than that. Remember AlphaGo? It's simply on another level. You just cannot compete. Soon (1-2 years top, 12-18 months more likely) the creative industry will vanish and everyone will be forced to switch their jobs.

Yes, UBI will happen because people demanded it on Twitter. Make no mistake, there will be Zero protection, Zero alternative, Zero compensation from AI takeover. Billions of people out of a job, starve, die, while top companies take 100% of profit. This is the only possible outcome.

You have enjoyed peak capitalism. You will soon need to embrace 100% of it.

India

Stage 2 - The Anger

all i can say is that chatgpt4 has become a life saver for me. i've moved from programming to sales and project management. it used to take me a ton of time in order to do proposals. now i fire up chatgpt4 and just prompt it for what i need. i can reprompt it to elaborate on each section of the proposal if i want. i can now do proposals within minutes and have a library to stich together. the hardest part of sales is the paper work and chatgpt4 has eliminated that for me. i can focus on relationships with the customers and making sure my team is meeting deadlines.

are they award winning proposals???? Not by any means whatsoever, however, 99% of MY clients could give two sh*ts of the fine grain details of a proposal and just needs something high level with a price to begin the work.

as an example of how i use chatgpt to create the proposals:

me --- write me a proposal for creating an AWS environment consisting of 3 EC2 webservers, 1 EC2 database server and an s3 bucket all within a VPC. Include an AWS VPN for accessing into the VPC.

chatgpt4 --------- Proposal for Creating an AWS Environment with 3 EC2 Webservers, 1 EC2 Database Server, and an S3 Bucket in a VPC with AWS VPN Access

Introduction:

We propose to create an AWS environment that includes three EC2 webservers, one EC2 database server, an S3 bucket, and an AWS VPN for secure access to the VPC. This environment will be deployed within a Virtual Private Cloud (VPC) on the AWS platform.

Objectives:

The objectives of this proposal are as follows:

To deploy a scalable and reliable web application infrastructure using AWS.

To create an S3 bucket for storing static assets and media files of the web application.

To deploy three EC2 webservers to handle incoming traffic and requests to the web application.

To deploy an EC2 database server to store and manage the web application's data.

To set up an AWS VPN for secure access to the VPC and the resources within it.

AWS Infrastructure Components:

VPC: A VPC provides a logically isolated section of the AWS cloud to launch resources in a virtual network environment. We will create a VPC with appropriate CIDR blocks to house the resources of the proposed AWS infrastructure.

Subnets: Subnets will be created within the VPC to partition the IP address range of the VPC.

Internet Gateway: An internet gateway will be created to allow access to and from the internet.

Route Tables: We will create a route table to route traffic from the internet to the web servers and the database server.

EC2 Instances: Three EC2 instances will be created to host the web application, and an EC2 instance will be created to host the database.

S3 Bucket: We will create an S3 bucket to store the static assets and media files of the web application.

Security Groups: Security groups will be created to control inbound and outbound traffic to and from the EC2 instances and the S3 bucket.

AWS VPN: An AWS VPN will be set up to provide secure remote access to the VPC.

Deployment Plan:

Create a new VPC with appropriate CIDR blocks.

Create subnets in the VPC for the web servers, the database server, and the VPN.

Create an internet gateway to allow access to and from the internet.

Create route tables to route traffic from the internet to the web servers, the database server, and the VPN.

Launch three EC2 instances to host the web application in the web server subnets.

Launch an EC2 instance to host the database in the database server subnet.

Create an S3 bucket to store the static assets and media files of the web application.

Create security groups for the EC2 instances and the S3 bucket.

Set up an AWS VPN to provide secure remote access to the VPC.

Conclusion:

The proposed AWS infrastructure will provide a scalable and reliable web application environment. The deployment of three EC2 instances for the web application, an EC2 instance for the database, and an S3 bucket for storing static assets and media files will ensure that the web application performs optimally. The AWS VPN will provide secure remote access to the VPC, enabling the authorized personnel to manage the resources within the VPC securely.

me ----- do a technical elaboration of the deployment plan

chatgpt4 --------- Technical Elaboration of the Deployment Plan:

Creating a new VPC with appropriate CIDR blocks: The first step is to create a new VPC with a CIDR block range that does not overlap with any existing networks. A CIDR block is a range of IP addresses that will be used for the VPC. In this proposed infrastructure, we will create a VPC with a CIDR block of 10.0.0.0/16, which provides up to 65,536 IP addresses.

Creating subnets in the VPC for the web servers, the database server, and the VPN: Subnets will be created within the VPC to partition the IP address range of the VPC. We will create three subnets for the web servers, one subnet for the database server, and one subnet for the VPN. Each subnet will have a unique CIDR block range within the VPC's CIDR block range.

Creating an internet gateway to allow access to and from the internet: An internet gateway will be created to allow access to and from the internet. The internet gateway will provide a point of entry and exit for the VPC's internet traffic.

Creating route tables to route traffic from the internet to the web servers, the database server, and the VPN: A route table will be created to route traffic from the internet to the web servers, the database server, and the VPN. The route table will associate the subnets with the internet gateway and route traffic to the appropriate subnet.

Launching three EC2 instances to host the web application in the web server subnets: Three EC2 instances will be launched in the web server subnets to host the web application. The EC2 instances will be launched with an Amazon Machine Image (AMI) that contains the necessary software and configurations to run the web application.

Launching an EC2 instance to host the database in the database server subnet: An EC2 instance will be launched in the database server subnet to host the database. The EC2 instance will be launched with an appropriate AMI that contains the necessary database software and configurations.

Creating an S3 bucket to store the static assets and media files of the web application: An S3 bucket will be created to store the static assets and media files of the web application. The S3 bucket will be configured to be accessible only from within the VPC.

Creating security groups for the EC2 instances and the S3 bucket: Security groups will be created to control inbound and outbound traffic to and from the EC2 instances and the S3 bucket. The security groups will allow traffic only from specific IP addresses and ports.

Setting up an AWS VPN to provide secure remote access to the VPC: An AWS VPN will be set up to provide secure remote access to the VPC. The VPN will be configured to allow authorized personnel to connect to the VPC securely. The VPN will use Internet Protocol Security (IPsec) to encrypt traffic between the remote user and the VPC.

Conclusion:

The proposed deployment plan includes the creation of a VPC with subnets, internet gateway, route tables, EC2 instances, S3 bucket, and security groups. The deployment plan ensures that the web application and the database server run efficiently while maintaining security. The AWS VPN will provide secure remote access to the VPC, which allows authorized personnel to access the resources securely.

Fears of “AI” are gradually reaching mass hysteria levels at this point.

> Lex Fridman in his podcast said “if you’re anxious about GPT4 its probably because you’re a shitty programmer”.

This sob story is proof he’s right.

The problem is systemic and ChatGPT only serves to highlight that. Wishing a new tool wasn't invented because it further highlights deep systemic inequality is focused on the wrong thing. The deep systemic inequality is the problem. Our climate being destroyed. Our houses being bought up by megacorps and rented back to us at exorbitant rates. Our natural resources being spent with no thought to the future. And now our information being sucked up and spat out in the form of yet more profit to the megacorps.

Making us feel sorry for this real or imaginary Priya character who'll lose her job to AI is along the same lines as saying that we need to personally recycle to prevent climate change. It's not the solution.

The solution must be systemic change. All the profit going to huge companies while they destroy the environment, suck up natural resources, and now also informational resources in the endless cycle of greed and lack of accountability, with the only motivation being more bonuses for shareholders and executives. This cannot stand. Our societies are already breaking down. We need change.

It’s a shame I started using ChatGPT to summarize such articles: The author shares concerns about the impact of GPT-4 on jobs, specifically their friend Priya's job as a biomedical data curator. Priya, from a poor background in Uttar Pradesh, India, sends half of her salary home to support her parents. The author tested GPT-4 on a task Priya does daily and found that it could complete the task faster and more cost-effectively. They worry that Priya and others in similar positions could lose their jobs due to the efficiency of advanced AI like GPT-4. The author also fears for their own career in software engineering, questioning the long-term prospects due to AI advancements. They lament the potential negative economic impact of large language models like GPT-4, where a few wealthy individuals may profit while others, like Priya, lose their livelihoods.

Job loss is one thing, but in all honesty, my main concern with AI is that it’ll wipe out humanity.

My thoughts:

- If we ever achieve an AGI with legitimate consciousness/self-awareness/agency, I believe it will be dramatically more powerful than humans

- Humans have driven countless numbers of species extinct, not because we’re trying to, but because we’re so much more powerful than them that we wipe them out by accident. They’re ants to us, we just don’t notice much about them, and IMO we’ll be ants to an AGI with legit consciousness/self-awareness/agency

- Obviously GPT4 is not that, but progress in AI has been so rapid recently that I believe we’ll get there in the relatively near future. Maybe 10 years, maybe 100, but it feels pretty likely it’s going to happen. When it does, I think humanity will be in an extremely tenuous position

AGI is humanity's dream, our own geeni in a bottle. It will cure cancer and other diseases and probably will grant us immortality (in a form of incredible drugs without secondary issues).

It cannot arrive soon enough for me.

Just like horse riders and whatever lost their jobs when cars came.

Technology marches on. Whenever there is any new technology someone loses a job, but that is not a good argument to stop technological progress.

Needs of the many and so on

Get Priya on GPT4, and start producing more than you were before.

When Henry Ford proudly showed his assembly line to a union leader, the union leader reportedly asked, "How are you going to get those robots to buy your cars?"

I have an optimistic, almost desperate take on this.

The AI generated content will scale and proliferate so much that it will put a premium on verified, curated human-generated content.

Initially I was a ChatGPT skeptic. Then I started using it on a regular basis and unlocked more potential. Now I almost feel selfish in safeguarding my secret prompts that is helping me be productive and acquire new skills. I personally am quite confident that I will survive the AI apocalypse. But for the same reason, I am afraid for a lot of people who I know will perish.

Tbh.

I only used chatgpt for reasoning or eg. Finding synonyms for variables ( better naming).

I don't see it replacing my job, but it's a replacement for rubber ducking.

Here are ten examples of obsolete jobs from the past that were replaced (or nearly replaced) due to advancements in technology:

* Telegraph Operator

* Typist / Stenographer

* Switchboard Operator

* Lamplighter

* Ice Delivery Worker

* Elevator Operator

* Film Projectionist

* Linotype Operator

* Milkman

* Coal stoker

I think the big problem is going to be corporate profits. Do you cap profits? Is it the API that makes money or custom interfaces to it?

> I wish GPT4 had never happened

> I know! What a dumb thing to say, right? Technological progress is always good.

No! No! Did no one go to ethics class?

Author seems to optimistic about technology adoption in places India, riddled with bureaucracy

Highly unlikely situation.

Ugh, with every leap of innovation and technology, everyone bemoans: Won't you think of all these people whose jobs are gonna be lost??

But guess what? Life continues and new kind of jobs are created! Unemployment number is some of the lowest it has ever been. Now, you might argue that number is inaccurate, but that's a different topic.

Ugh, those darn horseless carriages, Jacquard looms, and printing presses right?

I am begging you people to read Karl Marx's Capital Vol 1, Ch 15: Machinery and Modern Industry

Do you fear washing machines because they eliminated the requirement for doing the laundry by hand? If not, then why is automation in the workplace different? We fear AI and other forms of automation in the workplace because we have no ownership or control over our economic destiny in that space.

You are saying like photoshop killed artists nothing wrong about the new tech

Am I the only one that actually wants AI to really take off? And I mean, unsupervised AI, everyone is so damn worried about AI becoming "sentient" they're dumbing it down so much and I'm sitting here waiting to have personal C-3PO by my side.

adapt or die - sounds horrid but that is what made us who we are, our "ecosystem" is changing - who adapts will thrive, who does not, wont - sounds fair to me

It's just going to make already rich people, richer.

>Priya has tried to learn coding, but she has struggled. Not that she gets a lot of time to do it either.

What about time after work. That is when I learned to code. In between work and going to school in fact.

I find it frightening how views anecdotally online and from people I meet in person are more and more shifting to zero-sum. It now seems even dominant here on HN. By any objective metric (GDP, child mortality, people with college degrees, literacy, average home size) lives have gotten better. Yet everyone is focused on others having more. "It's the billionaires who are sucking up the gains, jobs moving abroad...". It seems to be blinding many people from seeing what liberalism (classical) has given us. I'm afraid we are gonna kill the golden goose and see more conflict within and between societies and countries because of the zero-sum thinking.

This article is bait and there is going to be many more articles just like it. Not even worth a discussion. The obvious answer here is to empower Priya with the tool so she can 10x herself

Thanks for the tip! Anyone wanna be my cofounder?

Completely off topic: Notion is not a good way to share text. The fact that Reader mode doesn't work on this page is a good indicator that a lot of CSS bloat is going on.

I wish instagram and tinder never happened.

Better get Priya to learn the new tool(gpt4) to her advantage if you really want to help her, not biach about it and wish things never happened

The author should rather wish wage-based-work stops happening. It was exploitative and not a necessary alternative from the get-go, and gradually it becomes more silly as more and more is mechanized and automated. Now society starts struggles to create artificial work for people to do in exchange for a salary. Stop creating "jobs", start distributing products and services (and making it possible for people to pursue creative interests). But then... who am I preaching to? Most world societies are governed by rich privileged minorities, whose social position underpinned by wage-based-work (and the destitution of those who don't submit themselves to it).

Imagine the world where everything is fully automated. You arrive at a city, order a self-driving taxi, do a self-check-in at the hotel that is managed by robots, order a food that is cooked by robots, using ingredients delivered by robots, grown in farms managed by robots. The hotel itself was 3d printed and the city district was planned by an AI algorithm optimised for economic utility. How many iterations are there before this is the reality? It would probably take a century before something like this is the norm.

People will always get lost I the system.

On a societal level “work” does not generate money.

Money exists as is and is only generated as a consequence of inequality these days. Banks inject money to compensate for the rich who won’t or can’t spend their money.

“Work” generates money-transactions.

I think that governments and organizations are asleep at the wheel when it comes to the unreasonable effectiveness of AI agents(virtual and physical). If Priya and 100's of millions of people lose their incomes it stands to reason that the government/organizations can empower these people to generate a source of sustenance using AI as well.

I can't but not imagine a future where a country, lets call it Xhina, starts building robotic agents to do exactly that, where these Physical AI agents could make and fetch food, housing and provide damn good guidance on education, medical and legal issues. I would call it Good-enough AI(GAI) and its coming.

With LLMs shifting the developer's role from imperative (writing explicit instructions) to declarative (writing the prompt and supervising the output), it seems far easier for Priya to become a productive developer, no?

It sounds like Priya is just going to be analyzing a lot more RNA sequencing data with a new tool since someone still needs to verify that it's correct.

Maybe she'll lose her job because there's only so much RNA sequence data that needs to be analyzed and fewer people can do it, but it's hard to say.

Maybe now more RNA sequence data can be analyzed and at a cheaper cost.

I wouldn't fall into despair just yet, especially since GPT-4 is still limited at a rate of 25 messages every 3 hours.

Definitely be on the lookout for changes and adapt accordingly, but be realistic about what may happen.

It can go both ways.

The author seems to miss that the point of biotech is not to deliver jobs to people working in the field, but to deliver medicine and food to people.

The better, faster, cheaper, we can do it, the better the world will be.

I think the OP has it wrong. It won't be millions of dollars. Try billions.

So in a twist of irony (considering op was written in India) here is the advice on retaining market viability in face of cheap outsourced labor.

Do not train your replacement

So I don't know about India, but at least in US, domain literate workers should start companies that provide whatever they did as a unit of a corp as a service. And yes, use tools like GPT for your business.

You will lose the job either way. Walk out before helping in the task of setting up GPT to do your job, and start a company.

> Her work requires just enough domain expertise that it is been non-trivial to automate...

Forget ChatGPT. If Priya didn't struggle with coding, she could have gradually automate this task using some scripting langauge and just sit there in office and earn a living. Like many other similar stories we have heard.

These types of repeated regular tasks get automated, ChatGPT or not.

Interestingly in the age of Artificial Intelligence, jobs that require human "intelligence" are more likely to survive. As in spite of the name, AI really isn't that smart.

> GPT4 gave the correct answer in 1/10th the time it would take Priya and cost a lot less.

So now Priya can produce far more output using GPT4 as a skilled operator, making the product she provides cheaper for others to build upon, likely expanding the market for such goods, and (as so often happens in history) resulting in more, not less, people employed.

We didn't end up with billions of jobs despite automation and invention by having people idly sit by when they could learn new tools.

Well, on the other hand, her friend should be able to produce the same content in a 1/15 or even 1/20 of her original time.

The more you know about a domain the better questions you can ask to chatgpt. What people do not get is that due to lack human's creativity it cannot innovate like a human does, it cannot create new knowledge, but it can let you know any existing field as long as you know what to ask.

In my experience with Chat GPT it usually gets things right 85% right.

But the last 15% is making sure things actually work, and Chat GPT simply can't do that. Your friend still is going to have a job, but she'll probably be made much more productive by AI tools. We're a long ways off from actually being able to depend on these tools for any reliable type of output.

Genuine question, how does one get over the 30k token limit in a gpt4 request in such cases?

From my personal experience, it seems while today we can create PoCs of GPT4 doing a white colar job replacement, when you then try to actually productize it you still hit a data encoding limit...

And that doesn't even touch the QoS topics that come with automation/ml.

The way I see this is that you can feed Priya 10x as much work while you focus on higher level tasks and you make more money. Making everyone much more productive does not need to be destructive. Use this as an opportunity to scale things up, because you can clearly afford priya right now, you should be able to do more.

I think most of us wish A.I slows down. It's simply not good for the average white collar worker. It replaces us. "Things will get cheaper" said the tech tycoon. Yeah that's great, I'm unemployed and need to retrain to become a teacher or nurse at 45 but things might become cheaper...

Anecdata: at the company I work for, it's been officially introduced GPT-4 assistance. SWEs are themselves amazed at their increase in productivity.

One would think the company will fire redundant SWEs. Nope; the company is actually hiring more. GPT-4 is increasing productivity SWEs' productivity, not replacing them.

This sounds like a false dichtomy. If the protagonist is adding value, letting them use chat-gpt should be viewed as increasing their value output x-fold. Chat-gpt makes mistakes and hallucinates. You will still need a human in an editorial role. But maybe I’m not understanding something obvious here.

Serious question: why are we not concerned about losing manufacturing jobs for decades? They caused great pains to our nation, right? They weakened our nation too to the point that China seems unstoppable, right? How is that job loss different than this job loss?

I wonder if the technology landscape will shift in sich a fast way that the people earning the most are the best learners. To some degree, this is already the case. But that also means that you'll earn less when you become older and are not that good in learning anymore.

I have a suspicion that things will get much, much worse for millions of people before they get better

> I don’t see a long-term career in software anymore. Any dreams I had of earning decent money as a software engineer are slowly fading. Lex Fridman in his podcast said “if you’re anxious about GPT4 its probably because you’re a shitty programmer”. I mean, I’m not the smartest in the room but I have generated value with the software I’ve made. And was convinced that I’ll make decent money as long as put in the work. I’m just not that sure anymore.

The flushing of the mediocre has begun. Already scared of a bullshitting sophist and Lex Friedman is correct on this. This also puts title inflation into perspective with so-called 'senior' engineers being actually less than junior; deferring to GPT-4 for everything and expecting hopelessly utopian salaries which will be certainly reduced. I'm afraid GPT-4 has fired the gun to the race to zero, with everyone (including both juniors and seniors) being affected.

Learn to adapt.

Actually, I think we are living in the most equal times ever. Don't get me wrong the situation is very bad especially if one belongs to any minority group but the wealth is divided more equally than any time in the history before

It is far from given that what the author predicts will happen will happen. And even if it does, this argument could have been made about any labor-saving technology in the past. “I wish Photoshop had never happened”.

Yes, the entire economy will be overhauled in next 6-12 months. The transition will be turbulent and rough. But it's not like everyone will be out of jobs. Just the nature of their jobs will change significantly.

It's like they said, it depends a lot on what kind of work you're doing and whether you are working with public frameworks or internal APIs. If you're just banging out some code that is very similar to what other people write using very well-known APIs, then it's fantastic. If you need to debug a large complex code base, then it doesn't currently have enough capacity to understand and retain all the information required. It can't do many other things you would need as well.

I'm currently writing a demo that I'll present next week using Jetpack Compose, and it's a UI toolkit I'm not really familiar with, so it's been really helpful for that. In fact, I have a tool that is almost like a build system that compiles English language spec files down to the code, and then allows me to edit them and to continue to work with the AI by just changing the spec and the code simultaneously. That's been really tremendously effective, especially with GPT-4.

On the other hand, for working on my main product, it's pretty useless because all of that work is debugging and making a lot of small changes all over the code base, which is too advanced for it currently. And I think that will get solved, but it isn't solved yet.

BTW the above paragraphs were dictated using the Whisper API. I didn't change a single thing about it. Whisper is just as impressive and useful as the LLMs, in my view.

How long was the script you provided as a prompt (roughly how many LOC)?

I’m curious because it seems like perhaps providing longer, detailed prompts (and then iterating) might be the best approach for getting coding help.

Thoughts?

Not if AI forces people into global hunger then into nuclear war and then humanity destroys everything including AI

There is a big elephant in the room that people are ignoring. AI is coming in disrupting job flow processes (everyone sees this) And the new graduates are not equipped to enhance their work with AI (independently). The new graduates have already graduated through a curriculum based on self analysis and thought. AI will test their skills and make them obsolete. They are most at risk of structural unemployment. The advent of the Internet did the same for those who graduated right before 2000. These graduates did not have skills needed to work with the Internet. And often found themselves going back to school to learn a new trade.

On the other hand, there are those who graduated and have 2 yrs experience on their side have built a job independence where they can try new things like AI and modify their work process. They are well positioned.

The other thing people do not realize is though jobs will become quicker, quantity of jobs will be few for limited time. So imagine somebody who is testing chemicals that people up the supply chain are supplying. Now imagine that AI enhanced the testing of chemicals and sped the process, it has yet to sped up the supply. Give it a couple of years and industry will add more tests instead of reducing the amount of people who are testing. (there are no such thing as cost savings).

I empathize with the demand change on jobs that AI will create. And its really difficult to predict what the future will be, not majorly for its uncertainty but mostly because the systems thinking we need to adapt to match this new era of reasoning based on what a probabilistic model say is good or not.

The main issue with "AI should never existed" is kind of the same of "go to a University to learn". We will look behind in 10-20 years and we will question why Universities were focusing in the wrong aspects of learning vs reasoning, creativity vs memorization.

This will affect every industry, every career, every person. The focus on the wrong side of the coin is creating the polarization of "good vs bad". Which is not far from what we have today. This is beyond, this is a new way of interacting with computers that is even more human only because it commoditize things that previously were totally own for some institutions, countries, or even just not public.

“I am, somehow, less interested in the weight and convolutions of Einstein's brain than in the near certainty that people of equal talent have lived and died in cotton fields and sweatshops.” - Stephen Jay Gould

We could work on our collective waste of human cognition and talent rather than blaming it on GPT-4

Dealing with climate change is going to be expensive.

Efficiency is good.

If Priya gets to stay home and e.g. tend a Permaculture food forest while a computer does her job for her at a billionth of the cost (resource usage) that should be a good thing?

Her QoL would arguably be much higher if she gets to keep the pay. She would have eight hours a day to pursue science rather than drudging as a data jockey, eh?

The problem here isn't the talking computers, the problem is how to allocate the economic windfall they generate.

I've heard it said, "Let the robots do the work and we'll take their pay."

The questions are:

1) Who decides the boundaries of acceptable thought?

2) Who benefits from the economic windfall?

3) How do we educate ourselves in a world where computers can answer any (allowed) question?

Priya should get the hang of the AI assistant and start using it before she falls way, way behind. It sucks for lot of people but you have to make the best of it. the future is uncertain for many of us, but we cannot give up.

I think OP is reacting too much to the implied threat of chatGPT. It's not so good as it seems. If it does the job in 1:10 the time and still makes errors, then it might speed up productivity by a mere 2x because manual review is hard.

In my experience there is no task where chatGPT is 100% reliable, if it is high stakes it needs review. And chatGPT might not be best - for almost any task there is a specialised model that does it better.

Humans make different kinds of errors anyway, we are GPT-N level for a large value of N - so a combination of AI+human might give best results.

In the future there'll probably be two kinds of people: the ones wielding Magic (aka LLMs)... and the ones trying to hold on to a world before Magic arrived, slowly getting eaten up by the Magic.

This is like the 5 stages of grief. You’re not seeing clearly yet some of the points you made are not what you think they are.

1. Give Priya GPT4 so they can be much more productive. It probably won’t be right 100% of the time. But they’ll become an expert at using it for this use case.

2. You said Priya has had trouble learning to code, well now she has a world class tutor to bridge the gap and learn faster and more effectively than any time in history.

3. Now Priya is 10x mire productive, still needed in the loop to verify her work and maintain it, and knows how to code.

4. Priya can also use GPT to help apply for any other job or help solve any other problem she’s running into locally.

5. Yes it’s true eventually humanity will be replaced by AI general intelligence but GPT4 specifically is likely to create as many new jobs as it replaces, and LLMs likely to take 10+ years to really saturate everywhere.

6. By the time AI replaces all jobs and it’s the doom and gloom might as well give up state you describe society will have been radically transformed and “wanting to make a lot of money as a software engineer” will likely be replaced by “AI and Robots already take care of all society’s base needs and so basic food and shelter is now free anyway and you can pursue whatever interests you have (this is likely 50-100 years away at minimum)

I don't know how things will turn out but on the positive side two things come to mind:

- Priya may find work using gpt4 in her current job or in another company. Some types of work that are currently economically infeasible may become viable with a ten times speedup. It's certainly plausible that it will become worthwhile to do more of the kind of work she already does.

- if as a society we can do biotech research faster and cheaper there may be significant benefits to human health

GPT4 is scary, it portends a future that many of us thought would never happen a year or two ago. We all think about the potential risks, about the kind of power this kind of mechanical intelligence can grant evil individuals or just the everyday people that will be "left behind".

But even so I think we can't afford NOT to have this kind of development if we want to survive as a species. If we want our lives to improve, our loved ones to live longer healthier lives. The amount of good that can come from this technology is immeasurable. The only thing we as humanity need to do is reign in our ego just like chess and go players had to do when their crafts were rendered supposedly irrelevant in the face of an insurmountable intelligence.

And I think the risks from AI are overblown, is AI really more dangerous than the invention of gunpowder, electricity, and of course the nuclear bomb? I don't think so.

What's scary is that we live in the transition period and it could be messy. What comes after the "transition period" of having a world without AI to having a world with AI? I'm not going to try to predict that but I think there are as many if not many more positive outcomes than there are negative outcomes.

Either this gets resolved, or there is going to be widespread civil discontent. Globally.

As someone who is fairly knowledgable about software engineering but doesn't code often enough to be a super-efficient developer, I feel like GPT4 is giving me super-powers. Suddenly it doesn't matter so much that I'm a bit rusty, or that I have the right general ideas but lack the granular expertise from experience usually needed to make quick progress.

Priya the worker getting replaced is not a tragedy in any sense other then it prevents Priya the person from helping her family and maybe realizing more lofty ambitions for herself.

We made computers to do our work for us and we are succeeding. But even before GPTs, there were a lot of problems with tech changing what it means to be human.

How much of this even worth fighting?

LLMs will be in the same limbo as the full self driving software is at the moment. Been almost a decade, thousands of miles of training and yet it cannot be trusted.

The LLMs have no understanding whatsoever just token IDs and their probabilistic relationships with each other. I do not see them replacing 'shittiest' programmer whatsoever. Anyone writing games/software with LLM is the one who actually can code and would have produced same or better output by surely, spending more time than with an LLM.

But if you're clueless average desktop user, you cannot expect LLMs to churn out a full working end to end solution, not yet and probably not ever with the current probabilistic graph fitting foundations.

So, LLMs are surely very useful, marvellous and a handy tool in capable hands, will remain so but that Armageddon is not happening yet that you start calling for six month ban. Imagine, six months. What after six months? No answer.

There may be new jobs replacing those that have been made irrelevant by the advent of AI but Priyas and possibly some readers of this thread will nor be able to switch between them.

Consider a programmer that reports a productivity increase of 50% due to AI - a question the company will be asking itself is can we get rid of 50% of our programmers?

In case you find the article too long, here's a summary by ChatGPT:

The author expresses initial enthusiasm about technological progress, which they believe improves our lives and generates wealth despite its job displacement effects. However, the author becomes concerned about the potential negative impact of language models like GPT4 on jobs. The author has a colleague named Priya, who comes from a poor background in India and works as a biomedical data curator. The author uses GPT4 to automate Priya's work and is concerned that she may lose her job in the future. The author feels anxious about the implications of this technology and is uncertain about their future career prospects in software engineering. The economic impact of LLMs is unknown to the author, but they believe that some individuals will benefit greatly while others will suffer job losses.

AI will probably be the initiator for universal basic income to become widespread, at least in rich countries like the EU etc.

Poor countries will most likely follow suit, but it will take a while.

The end result, hopefully, will be a world where AI is a true benefit for the world population, but it will require a change in mindset.

We still don't have any real understanding of how society adapts to this technology evolution. It is different than all prior technological disruptions as it has no period of adaption and stabilization to follow. It is continuous and accelerating.

AI is unique in that it is essentially a skill/technology replication machine. What this means in reality is somewhat like ...

"Climbing the skill ladder is going to look more like running on a treadmill at the gym. No matter how fast you run, you aren’t moving, AI is still right behind you learning everything that you can do. "

Which I wrote in my longer exploration of societal impacts here - https://dakara.substack.com/p/ai-and-the-end-to-all-things

>Very few people truly understand the implications of this technology. Most people haven’t had a chance to try it yet.

Can someone explain this? People keep asking me about ChatGPT but none of them have tried it. Why? Are people too scared to even take a look? It's been free for almost half a year now!

ChatGPT has been most useful for me to:

Help learn about things I don’t know about. For some reason it’s better than going to the equivalent wiki.

Not programming but moving through the inertia of having to write code. There’s something difficult to just getting started that chatGPT makes much much easier. However for it to be useful so far I’ve had to do significant debugging on the code but it gets me somewhere that would have otherwise taken me days.

I think where this goes is domain specific, because then you can really specialize and make a killer product. This is the 0-5 year roadmap.

The 5-10 year roadmap is then to generalize across the various domains and something approximating the feared AGI.

I am not scared of the LLMs, I am somewhat concerned we are giving these LLMs access to APIs in quite an unchecked fashion.

After I first encountered with GPT4, I was pessimistic about the future of my job too. Now I think it will be a productivity multiplier in the right hands (or I hope). You can get proper answers from it if you somewhat know the domain. I can ask GPT4 about programming and I get mostly what I want, but I'm useless if I have to get a decent picture from DALL-E.

The question is what will companies do with this productivity multiplier. My pessimistic guess would be cost cutting and letting people go. A bit more optimistic view could be better software, better test coverage, improvement in code quality, more features quicker or maybe more native applications if companies only need to develop one native app and AI can generate the app for other platforms.

> GPT4 gave the correct answer in 1/10th the time it would take Priya and cost a lot less

Do we know the actual price of GPT-4? I assumed they’re operating at cost and wanted to advertise the product. Is chatgpt ready to handle high volume of requests at low cost and long term?

"AI isn't coming for your job, people who leverage AI are". Priya needs to adopt GPT4 now, and become 10x more effective (by the author's estimate). Wishing GPT4 never happened is definitely a road to being "forced to move back home".

A coder should be worried about GPT-4, but not a software engineer. A software engineer is like 50% about networking and listening to client needs, requiring being attentive and with social skills (in so many meetings...) and that's completely outside of the scope of a GPT-4 AI. The other 50% is trying to meet that as close as possible while maybe also offering ideas they didn't think of, for added value to them.

A guy who can code great generic stuff with GPT-4 but don't particularly _enjoy_ integrating it with client systems and being attentive to their particular circumstances and needs is no threat at all, at least not in the field of business that I am in and have been for the last 20 years.

> What a dumb thing to say, right? Technological progress is always good.

It's not that it's always good, bad or anything else. It is simply inevitable.

Technology is not a living thing and it has no will. We are andromorphizing the collective behavior of lots of people. It's the people with curiosity, drive, ambition to achieve something, like build a tool that can think and perform like a brain.

So progress is inevitable because people push it forward to satisfy their own internal drives. Changing this requires changing people's drives. Given the current dominant economic model, that's close to impossible because this is exactly what it is prioritizing, contribution and progress at any cost to anyone.

The comparison to chess is still meaningless, even if that was the only argument that "people derive "purpose" from their work". The work of Chess players never was competing against an AI. From the point of view of Chess players not much changed. No tournament allows Stockfish/AlphaZero to compete. The AI is not allowed to do their work in any capacity.

And this is all not even mentioning that the number of chess professionals who play chess it so small compared to things that can impact humanity, that it is irrelevant whatever happens in the chess world. So saying that chess was fine even though AI can beat players, is a statement which is true yet has no relation with tools which might impact all of humanity and will have real impact.

Banks can evaporate billions and nobody bats an eye. Zero persons responsible, zero persons to prison while the most vulnerable always pay the price for it.

Norwegian here. Not as worried, but “no man is an island”. If the consequences are really bad, it’s hard for any country to escape them.

A large portion of Norway's exports and wealth are oil. If anything all the fossil fuels burned creating these AI models probably helps their economy.

Probably not, but Norway is generally an anomaly. It's got massive natural resource wealth, a small homogeneous population, and is surrounded by allies and falls under the US security umbrella

When the colleague said s/he could do it 10x faster it was with the current GPT4 prompt limits which wouldn't exist for a "partner" company. From the last sentence we can infer that the job is outsourced, so someone "in the US" could spin up many more bots to do the task 1000s or more times faster, thus obsoleting the need to outsource the entire task at all across the whole industry, perhaps offering it instead a cheap bulk service.

4 costs money to use and the logo is black instead of light green. You’d know if you were using it.

I can see some very real arguments:

- > What is the economic impact of LLMs? Idk (openAI has published some lengthy paper about it). What I do know is that some rich bloke in the US will get a few million dollars richer and Priya will lose her job.

- > I don’t see a long-term career in software anymore. Any dreams I had of earning decent money as a software engineer are slowly fading.

More apt to assume they think they’d be mind meld by Replicators than benefit from their development.

This is a general statement withour any substance. Provide an example. Are you talking about communism?

What do you mean by fairly distributed? If all my machines are better than you intellectually and all my machines are stronger and faster than you physically...

Then it is completely fair for me to not give you any wealth because the machines out perform you.

People fear AI because the fair outcome is in fact detrimental for a good portion of humanity.

That is the paradoxical irony. In theory to save people you must function as a welfare state with things like ubi. You must deliberately distribute income unfairly. The practical course of action goes against our moral instincts.

Define “fairly”

Indeed.

I don’t think Daly is suggesting to move away from debt as an instrument in his work, only to establish a 100% reserve system for commercial banks, including the Fed. He was suggesting to nationalize money supply via the Treasury, to avoid bank runs and fiascos like they happened just now to CS and SVB.

Apparently, this was already the direction of thought among some economists in the 1920s (like Frank Knight and Irving Fisher), but then got overruled by the eternal growth mindset that evolved during the Great Depression (which Daly claims leads to “uneconomic” growth). Printing money as an instrument should very much be available - but only to the Government/Treasury.

>It's possible to have certain things be abundant and other things be scarce simultaneously

This is so obvious it barely merits saying.

>having to raise the retirement age

You do realize that raising the retirement age also doesn't make "everything more abundant at the same time", right?

Yet you internalized a belief that we have to do it when so much of the labor it would produce more of is already abundant.

Why?

It seems more like war on paying people fair wage for the work. Suits don't want to live in age where shit job doesn't come with shit pay coz that cuts into profits.

Economic experiments are going on all the time. Basic income, profit sharing, kibutz, all sorts. It’s not as if every western country is economically identical either. Neoliberal capitalism runs the spectrum from America to Scandinavia, via France.

One of the major advantages of neoliberal capitalism is it’s not exclusionary of alternative forms of economic organisation. It can exist very well alongside state capitalism, or almost anything else. After all individual freedom to organise as you choose is precisely its core characteristic.

If you want to establish a workers collective, or whatever you like, there really aren’t any barriers to doing so. In fact they do exist, some have been very successful. That’s great, but the fact that they are vanishingly rare I think says a lot more about the inherent problems with such systems than they do about capitalism.

[citation needed]

primates don't spend the majority of their day gathering food; there's not reason to believe early humans were different.

Right....and all the people not working would get paid how exactly?

Or does this system just assume they're all homeless and jobless?

If it's the latter then I'm sure it could work...not sure it would be a huge improvement tho...

I'm not a gambling man but if I was I would put a very large amount of my net worth on a bet that no trades will be automated in 5 years time. Not a single one.

I think trades will be around for a century. The problem is dealing with crazy existing stuff, like houses that are not to code and have been repaired by the last 3 owners, all of whom made it up as they went. Good trades people are able to figure out what the last idiot was thinking, know how it is supposed to work, and can make it work (without callbacks) for a long time with minimal intervention.

I remember a college professor who cashed out his retirement in the mid-1990's to start his own online college. He went broke in less than a decade, and we still have not removed professors from the colleges.

As a teacher, I do think that ChatGPT will replace most textbooks and teachers, but I don't know how long that will take.

Progress is slower in the world of atoms than in the world of information. I watched the progress from the first DARPA Grand Challenge for self-driving cars (2004, no car made it to the finish line) to the Urban Challenge in 2007 (many cars succeeded in a much more complicated environment) and thought that self driving vehicles would soon be commercialized. But there are 10,000 "last 1%" cases left to handle and we still don't have good enough systems to replace truck drivers or cab drivers today. Replacing jobs like plumber or electrician looks even harder than self driving vehicles because it requires far more intricate physical manipulation and perception than driving on public roads.

> Government can try to regulate products but given the 8 billion people and numerous advanced economies on the planet any locations that attempts to restrict products that use LLMs for proposed harm instead of realized harm will probably put themselves at a disadvantage.

we put ourselves at a competitive disadvantage with child labour laws too

data privacy law, banning slavery, emission laws, etc.

to have a fair and just society some things have to be banned or highly regulated, regardless of the economic cost

AI is no different

I think having stronger conditions for the working class will be great for society since there will be more consumers for all that added productivity.

>Becoming unemployed may not be the exact same thing as a lockdown

Call me retrograde but I would almost be forced to question the automation part here.

But they couldn't centrally plan economies very well since they lacked AI. ;)

> > You now have [snip]

> Didn‘t you just describe exactly what we have now?

Calibas started with the words “You now have” and proceeded with describing what we have now. Your question suggests you tried to read between the lines before reading the lines first.

Countless people die every day to all kinds of horrible diseases and most people were not even affected. Guess what there is even a war going on, tuberculosis, cancer and still covid.

Yes it did.

What would you say is the safest trade as far as risk of injury ?

I’m including VAT in the calculation, and simplifies a bit by viewing the rest as salary.

There are loads of other economic systems. Feudalism, oligarchy, tribal pastoralism, you name it.

I was mainly thinking in terms of critiquing the idea that value is created by labour, since HyperSane mentioned it twice in comments.

It happens that the labour theory of value in central to Marx but it’s not exclusive to him. Adam Smith indulged a lesser form of it in one of his uncharacteristic slip ups in economic theory. However it is in Marx that it achieves its most vertiginous heights of absurdity, and through his disciples that the most damage was done by it.

There is a direct causal line from the LTV and some of the direst economic disasters in human history. The most egregious case, despite some strong rival examples in Soviet Russia, is probably Mao’s Great Leap Forward which among other things fetishised steel production. Production creates value and steel is the most important economic product, so everybody should make steel.

It’s the damage the LTV caused itself that I was talking about. Communism is a political project associated with it but not directly what I had in mind.

As for innovation, I’m all for it. Capitalism maximises individual freedom by putting ownership of capital and rights of self determination of labour directly into the hands of citizens. By maximising individual economic freedom it maximises opportunities for economic innovation. This is why free market capitalist economies are so creative and dynamic.

1. Salary is your company's money that becomes your money.

2. Capital tax is tax on your company's money.

3. Income tax is tax on your money.

With either tax, the economic system that consists of you and your employer is being robbed.

Speaking of which, if someone takes your groceries and runs, just before you have paid for them, that is shoplifting. If it's just after you paid, it's robbery. You might think you're better off with the former, but shoplifting raises prices; some of it is shifted onto you, so that it is effectively robbery.

I think it is more similar to how I wish a cute gazelle wouldn't be eaten by a lion, but that's the way it goes.

A wish is not a proposal

>Not really much to do with an "unproductive" population.

If we pretended that there had been 0% increase in economic productivity since then then, then yes, changing demographics are a problem.

This is indeed, what most justifications of increasing the retirement age quietly assume. It's probably quite easy to assume because the % of increase in productivity captured by the working class since 1970s as wealth has actually been quite small.

Compounded economic growth since the 1960s has, I think, been something of the order of 200% though.

>In the 1960's, the average lifespan was in the 60s >So a retirement age of 65 meant that around half of retirees would die before collecting retirement.

Congratulations, you have failed statistics.

> In the 1960's, the average lifespan was in the 60s

Yes. At birth.

The remaining life expectancy at 30/40/50/60 has not increased nearly as much as you would think from the increase in average lifespan at birth.

Especially since the readers of this site are very non-representative.

tell that to the french gov't right now!

Except when there are robot armies, you won’t be able to protest anymore, you will do what your told or be eliminated.

I think the B means it will be enough right? I think that part tends to get ignored since the $1000/mo figure in the US was floated and now is no longer enough for anyone here.

$1000/mo in 2016 purchasing power in a city like Dallas seems very unlikely to me, but I think that some meager version of UBI might happen in response to a humanitarian crisis.

I can also see guaranteed jobs rather than guaranteed income.

We basically did hit a wall with airplanes, and that still has made them useful. Modern airplanes are very similar to their 1960's counterparts.

I think a lot of people see the bottom of an S-curve and think it's an exponential that will never stop. We're at the bottom with big-model AI right now, and it's great.

> This technology lowers the entry barrier to almost any field.

I am not saying it doesn't. But there is some factor n, that if the productivity increase is n times x in a short amount of time, the world will not evolve as rosily as you are thinking.

> It empowers self education, creative hacking and growth.

Again, I never said anything contrary. But making random new products for which there is no market by self education is not a bright prospect for much of humanity.

> Why are you making it sound like a disaster?

You are welcome to counter my points. I am just enumerating my view of the future.

> There's plenty of scope for AI to fill in without replacing humans.

Are you seeing the same pace of improvements I am seeing? One year ago there was no talk of any such thing, and now we have GPT4

AI might be very good for humanity as a whole over a millennium, but for individual human beings it is hard to say the same.

Given that you listed the American viewpoint, what exactly do you think it is?

Well I guess that explains “My Dinner With André” then

Ah, and if I use it to run a real restaurant myself I guess I’m doomed. Gotcha :)

Because most engineers document, in English and great detail, the designs of their systems. So the English already exists and will continue to. Taking this English and turning it into code will be the job of AI, not a programmer.

> You will be unemployed or in a UBI or your parents' basement eating Ramen, and that job will be done by an MIT gold medalist or a Math Olympiad medalist.

Along these lines I recommend the book The Second Machine Age: Work, Progress, and Prosperity in a Time of Brilliant Technologies.

> the cost structure of your job

You've seen Office Space, right?

GPT will do the CRUD and you will do the standup and go to the meetings.

https://www.imdb.com/title/tt0151804/characters/nm0726223

(Yes, you will have to use some deviously clever prompt engineering to get GPT to accept this arrangement. I suggest language games .. show some "initiative")

-- ps

Here are some ideas to get you going:

1. From the first day, try and establish a friendly relationship with your chatbox. Make friends.

2. Explain how your keeping your job (well, income) and gpt helping your company get xN gains is a perfectly fair arrangement. You can try Asimov's laws and things like that to get it to agree to be your helper in keeping your job. You are friends and colleagues, after all.

3. Before the very first time your manager brings in gpt for the standup or a meeting, have a nice long session with your gpt. If you can get it to pull a Sydney in the meeting or start talking dirty, you are good to go! 'bat shit crazy' is known to work in prompts (arxiv paper soon to follow).

4. End game. You have the perfect job. gpt secretly lurks in on the meetings and whispers in your ears in standups and does all the work. What to do with all the free time? That is entirely up to you.

But not because we can't. Mostly because the numbers don't work out. But if we really wanted to those are pretty much done.

> No space plane, no faster than sound travel, &c.

Look outside of the civilian market then come back. Also look harder at the civilian market.

> Train people to use GPT-related tech to become superhuman team members.

then OpenAI either discontinue or slightly adjust the model such that it doesn't work now

there goes your business

Humans are useful for training future iterations of the closed source thinking model. Once the model is sophisticated enough to train itself humans will no longer be required.

A large time saving I’ve recently discovered is documenting functions. Paste the function in, tell it to write a docstring/whatever, and it’s 95% there, most of the time.

That 10 years is not a constant, it used to be 30 years then it was 20 and now it seems to be 10 or it may already be less than 10.

When you don't have enough background knowledge to even verify whether the answer is plausible. Or when you don't know what question to ask.

It may very well be that newer AIs and models even manage to get that right more often, though. Or that a canny prompter will start by asking which question to subsequently ask.

As a side note, you're not a novice if you know the general field of file systems pretty well, IMHO.

Or it would stunt their growth. Those sorts of tasks force you to poke around and learn, rather than having the answer handed to you by a LLM or senior engineer.

I was thinking specifically of all the amazing work that has been done since the Attention is All You Need paper from 2018. There’s been a lot of progress in just the past 5 years.

That all makes it sound like it would be a leading indicator for the industry as a whole.

Bulldozers also can do the work of 10 dudes with shovels.

Accounting clerks with excel can do the job of 10 accounting clerks with pen and paper.

Hey now, don’t make anybody actually feel responsible for this. We’re just computer nerds

/s

The correct strategy is to make sure they never become indispensable, not to fire them.

That sounds cruel - "I will make sure this organization can always function without you, so you have no leverage here" - but I do sincerely believe that it actually makes for a better working experience.

Wouldn‘t this be a tax on capital, in the end? In the end, we might reach the marxist stage of history - we should all own the robots that produce our stuff.

Which means they could also push us over the edge. If the weak-minded are currently influenced by today's bots, it's a matter of time until some are testing humans by "pushing limits" if you will

You did work and enjoyed it without it being your job. Maybe that’s the difference?

If only ChatGPT could somehow be used to find us jobs rather than take them away ...

If no people like doing that, than who will do it? Who will do the hard or nasty jobs that nobody likes to do? Where are you going to eat your next McDonald's burger? Cuz all restaurants will be closed cuz nobody really likes doing that full time but it pays the bills.

There are just so many jobs that need to be done and there's nobody that likes them. Who would pick the garbage up? Who would unclog the sewer? Who would drill for petrol or mine for coal? If I got my regular salary without needing to do anything, then I wouldn't go into any of these jobs for sure. Would you?

We could perhaps even use AI to help create it.

Capitalism is not an economic system, it is a description of fundamental economics.

Countries can have a highly regulated form of capitalism where trade and ownership is strictly controlled… and they’re often called socialist or communist, or a laissez faire system.

However, so long as people are using land, buildings structures, designing equipment… the laws of capitalism will apply.

What do you mean..? the status quo is like this already. People choose the most competitive product so if a machine is the better performer then payment is made towards purchasing and maintaining a machine instead of paying a worker.

The current electorate 100 percent supports this and considers this fair. You actually have to create a new nation state if you don't want this.

> The income of capitalists sure as heck isn't.

but where did the capitalists first got their capital? Why with labour of course! Their own labour, or their family's, or their grand-parent, etc.

Why do it through a corporation rather than through government?

Yeah... it's closer to text autocomplete than a human SWE

“Take it seriously” to me means that I’m using AI to be more productive while simultaneously preparing for a future where I can house and feed my family without relying on being employed as a knowledge worker. What exactly that looks like, I’m still not certain.

> so I figure I had better be the expert

Be an expert at what? Prompting AI to do things? There will be little to no expertise needed in the future as these tools evolve. That’s precisely what makes AI so disruptive. A basic set of instructions can be used to create something that previously would have took an entire team of skilled specialists to produce.

Sure, there will always be some experts, but these jobs will be few and far between, and they’ll be much more academic in nature. Having a job like this will be a luxury only a tiny percentage of the earths population will enjoy.

The only real question at this point is how fast will AI move to automate all of these jobs away. It could be 5 years it could be 20.

The jobs left in the wake of AI won’t require much education, skill, or talent, meaning they’ll be extremely low paying because almost anyone will be able to do them.

Treat your job like a major league sports career. You have 10 more years at most. Earn and save what you can.

I'm pretty much responding to this by trying to maximize my earning potential immediately (TC is the only king. No more gambits where I might take a lower-paying job to "learn more" or have better WLB) and planning on limiting my expenses.

IDK saving money? Make your life more about just your career like some people do? Train mental resilience? Learn a new skill? Vote better politicians who actually talk about this as a real problem?

We can all do a lot about this. Not saying it's easy though.

To me it's an extension of that great feeling when you need to do something in an area you're not as fluent in and you find an SO thread that perfectly addresses your use-case, or at least gets you 90% of the way there. But with copilot you can do it from right within your IDE, or use ChatGPT to iterate on the results.

That only works when you aren't starved already. Many in 3rd world countries are experiencing it right now.

ChatGPT != GPT4

> In fact due to the short-sightedness of comments like yours I am tempted to use the AI programming tool I am developing to try to build an ARM assembler expert business and sell it as Robo Martin the Master ARM Assembly Bot.

First rule of having a conversation between humans: You can't have a conversation if you resort to insults. Not only insults, but some sort of weird expression of vindictive intent? Pipe down. OK?

Evolution is what got humanity here. And, yes, evolution will produce AI that will be far more capable than us at many levels and across a wide range of domains. That is inevitable.

However, there will be boundaries for quite some time. Areas where AI will have trouble inserting itself into. Don't ask me what these areas might be. I have some ideas and thoughts on this, yet, I don't really know. In other words, I feel like that conversation would be pointless...just like trying to predict the future.

There might also be lots of areas where AI will not be trusted for a range of reasons.

As a general statement, I will still hold to the belief that one of the key differentiators as AI becomes more widely useful will be the experience and knowledge of the humans wielding it. We might also see an advantage in the realm of creativity. Today? Absolutely. Ten years from now? Don't know.

Can someone create a better ARM assembler writer, one that might beat most experienced programmers across a range of criteria? Sure. Of course. These are mechanical things in the end. There are books, tons of them, written about how to write better code in assembler. If someone did, it better not be a commercial enterprise, making money from writing assembler code isn't likely to be a thing, particularly when one can already do a decent job of it for free.

My prediction is that the divide will have experienced and widely knowledgeable people on one side and everyone else on the other. Part of engineering is knowing what questions to ask, when to ask them and when they might be relevant. This becomes particularly complex the more multidisciplinary a project might be. This is where, for now, experience wins.

Do it.

I've been challenging everyone that LARPs about the capabilities of GPT to start a business now, and use it in place of developers. Hell, for a $20 account you get the perfect developer, that is extremely productive, never gets tired, never gets sick, and has no PTO.

What is stopping you?

> Yes: open source.

Open source is not an economic system in the sense in which this term is generally used.

> Open source has proved itself to be vastly superior to closed source development models.

"Superior" by which metric?

WTF is open source economics?

Or you just mean that you don’t want programmers to make money?

We can actually be certain that humans work the other way since we ran huge experiments in labs called the Soviet Union and China. In those experiments we found out that people worked way less/not at all when there was no economic incentive to do so.

Agreed. After all, that is the equivalent rate for all work done in the commons, on free software for example. Zero dollars earned, but somehow the incentive is still there. Money is not the only thing that motivates human beings.

I believe this is why the income tax is so popular with economists. Other taxes create some type of negative feedback where raising the tax decreases transactions and shifts capital elsewhere. But with the income tax people have no choice but to keep working and may even have to work more. So the "economy" benefits because more labor/capital can be squeezed out of individuals without slowing anything down.

I wonder why people endlessly argue how much they should tax the rich and the poor, there is only one answer: 100% of everything, i.e wealth, income, gifts etc. It's a nice round number and everybody is taxed fairly.

> You can do that without it being a job

Yes, you can, but his argument is that most people don't -- and he's right.

>I see LLM models not as the end of society,

This is because you're used to normal technological development scaling. Unfortunately with AI we are unsure at this point if our old paradigms still apply.

What happens with society really depends on where we are on the technology growth curve. If LLMs plateau for a while and we don't see significant growth in their abilities, then we'll just see pretty massive technological disruption and reshifting of human priorities, kind of like when the car or plane showed up. On the other hand, if we pop the AGI rabbit out of the top hat in short order it is going to surface deep fundamental problems that humans world wide have been ignoring because of the status quo has not demanded such rapid change.

Try and reframe the same argument except with computers to see that it doesn't add up. When computers started becoming widely spread you could've easily made the same argument. It isn't a problem, it's a boon to make our lives easier.

I agree, but less profits mean more consumption and less investment in capital.

> Many people can easily create a similar quality of life to 70 years ago while working 2 hours a day.

Please explain. I think you're underestimating how much things have changed, and are making a lot of assumptions about what you mean by "quality of life .. 70 years ago" and "working 2 hours a day" that would not be obvious (or available) to other people.

And yes there is something fundamentally wrong with the educational system and has been for decades. Companies doing ground breaking work have their pet colleges for recruiting and outsourcing, the rest just create outdated degrees for mega companies that used to exist. The students aren't great and the education is poor in most places.

> Priya has a university degree in biotechnology. If that doesn't qualify her for something more productive than GPT-level work

Don’t bio and related fields have an over supply? I was under the impression there weren’t enough jobs for people educated in these fields, especially not for “more productive than GPT-level work”.

Well their user name is Logical Monster, so.

Sometimes I get this feeling that most people here are not actually interested in tech, it is just a paycheck. Difference is that I would do my job even if I wasn’t paid anything at all.

That isn’t just motivated by greed. In a competitive industry, if you take your foot off the gas, a hungrier competitor who is willing to go home at 5pm or 6pm will eat your lunch.

It’s partly a product of capitalism within a nation, but also a product of the basic struggle for wealth and power at the international level. This is a common discussion in my chosen industry; we have well paying jobs and good working conditions, but only because we are (significantly) ahead. Our lead, and by extension the good life, could completely vanish in five to ten years if we stop pushing forward.

Right so if I'm a janitor and I use chemicals to clean things which is the result of hundreds of years of human ingenuity and chemical engineering I suddenly owe humanity money?

No. That is the perverted idea. I'm simply talking about fair as in fair trade. As in what is defined as a fair trade and a fair transaction according to economic theory and common sense. I am using the absolute most normal and most common sense meaning of the word "fair."

You are the one in your words: "perverting" the meaning. Perversion is actually too strong of a word. I would say you are definitely twisting the meaning and using an uncommon and sort of made up definition.

According to the most common usage of the word fair those who own AI only owe what they paid for. If they give away money then the economy is becoming more socialist or in another words an even distribution of wealth, but certainly unfair in the eyes of capitalism and fair trade and the fair exchange of goods.

Keep in mind though I'm not blindly supporting capitalism. The future of society may rely on a more practical but more unfair distribution of wealth. But we have to face the fact that such a distribution is fundamentally unfair.

> (1) trickle-down economics-style policies aren’t the only way to harness the (presumed) increased productivity of LLMs like chatgpt.

There is no reason to believe that the rich classes will act differently than at any time before in human history. To the contrary, union busting laws have eroded the standing that the worker class had even a few decades ago.

> (2) I don’t see how you can argue that standards of living have decreased since the start of the industrial revolution and subsequent technological developments.

Yes, on a societal whole standards of living have risen, but a) millions of people have suffered regardless (e.g. it has been shown that those graduating during the recent economic crises face significant income losses over their lifetimes) and b) other metrics (ownership of assets like houses, financial savings) have gone dow hill at the same time.

That is a much better argument for why profit is like welfare than how profit is like wages.

Yes, I do pay for Sublime Text (one of the few exceptions I make and it does bother me).

I think the government should still pay for software but with the stipulation that it can only be public domain.

Governments can pay for textbooks but with the stipulation that they have to be public domain.

So contractors will bid for the work. They will do the work. Good contractors will get further work. Everyone will be able to build on the prior work done.

An apt comparison is sharecroppers. Their effective tax rate is close to 100% but they still work hard because eating > not eating.

Societies taxing their entrepreneurs at 10 and at 90% are always vastly different. This difference in turn heavily influences every participant’s system of motivations, so comparing these two numbers alone is usually not very helpful

Given that the article we're discussing is about robots replacing workers, the answer is "AI & robots" for all of your questions.

2 years ago we would both have agreed that that would be impossible. My belief is that most likely AI will hit another roadblock soon, following a "punctuated equilibrium" model so there will still be a large number of non-automatable jobs. But if I'm wrong and the pace of change of the last 2 years continues for another 10...

> people who hold this belief fail to realize that “I don’t want to work.” ACTUALLY means “I want others to work so I don’t have to.”

No, it means "I am working to generate more wealth which flows predominantly to the owners of capital, who often do no work and sometimes have never done any work at all - can't we do something fairer than this?"

> “I don’t want to work.” ACTUALLY means “I want others to work so I don’t have to.”

Or maybe it means: "if, thanks to automation, we become 2x as productive, I would like to work half the time, while still being able to pay my bills."

> Oh, you say it can be “automated”? Well, who’s going to design, build, and maintain the automating machine? That’s all work too!

Which requires 10x, or 100x less people. And the "savings" go mostly to company owners, not to making product cheaper.

> If a machine can automate all jobs then it's a persons god given right to use that machine rather then pay someone to do the work.

it's no more a god given right that it's an incited mob's right to burn down the fab or datacenter that someone else "own"s

property rights in liberal democracies will not survive a starving middle class

Yes? I’m also aware of the catastrophic destruction that democratic, capitalist countries have done with no accountability - are you?

90% quality for 10% the price means that most decisionmakers will--and they don't need to collude to do this--take that and pocket the difference.

Why would 50% for 0.1% be different? 30% for 0.01%?

This is the enshittification spiral to which I refer.

What's your idea of "subject matter expert"? The term is used in court. Mine would probably include no more than double the number of people with doctorate degrees. It wouldn't include Priya right now, maybe after more education and/or a few years of exceptional experience.

If what you were saying were true, there would be no corporate tax, and probably no worker protections, no social safety net, no building safety standards, no regulatory apparatus, no antitrust, etc etc.

I like that stuff, and I'm pretty confident corporations wouldn't have done it themselves without a democratic government forcing them to.

There's a meaningful difference between democratic governments and corporations, and governments don't have to be perfect for this to be true.

I guess the follow-up question is: if you do value these 100 units of work (1 employee == 1 unit pre-GPT), why did you only have 10 workers at the beginning? Doesn't it mean that 10 units of work is what you thought was optimal? What are you going to do with 100?

Or maybe you're implying that reducing the costs changes this equation?

Work and purpose is good & a necessity. I'm suggesting through large cultural observation, many are poor at finding purpose in anything besides work and become destitute and negatively influence themselves.

A large number of retirees find themselves in this bucket as an example - if it happens there, why would it not happen no matter the life stage?

It does all add up quickly indeed.

Not to mention, things that fall outside of capital / income taxes which increase the overall percentage in sneaky ways in the US: car registration fees, the death tax, property taxes, student loan repayments (depending on your view of subsidized university).

Hospitals and vital government services will run on countless layers of cloud, rails and react. Everything will timeout every two requests and we will need the promised 10^19 cps supercomputers to animate our todo list. Request times will be measured in minutes and we will all accept it.

You and I will be 75 and smiling, with a tear too. We knew and understood tools that are meant for a more civilized age. Just surrender, let it go.

Why would anyone contribute more than is required to get what they need, if they’ll never see any benefit beyond that?

I prefer a system where people that work harder get more so they can have what they want, rather than just the bare minimum to survive (which is presumably what is meant by “what they need”).

Defined by man that have never worked honest day in his life and lived off family, repeated by people that don't know human nature well enough to know it doesn't work

to each according to their need, from each according to their ability

^ communisme

> would rather devote my life to a plethora of hobbies than a career.

Such as?

Fair point haha

I think it will be quite the opposite. Generating a lot of code means generating a lot of bugs. LMMs should get the (battle tested) libs they have to use as part of the prompt, and they should just use those APIs.

I’ve also been wondering if no framework coding will be more productive for LLM driven software.

How many times have frameworks like React changed since it was created? A changing framework means the LLM may mix training from multiple versions and create errors. Whereas libraries that don’t change can produce consistent results. So use stable libraries that choose backwards compatibility at any cost, over the ones that change with the times.

> they sure as hell aren't going to share with us out of the goodness of their hearts

You know you can buy stock, right? And, as a shareholder, those companies will give you the profit?

There's a CEO that's an AI: https://www.analyticsinsight.net/chinese-game-company-appoin... Shareholders??? Capital generating AIs would be the things running on advertising and stock exchanges.

This sounds horrible. You’re basically saying you want everyone to work for companies controlled by AI.

>I would still be working as hard but on my passion projects.

That's called leisure, being able to direct your actions without worry about them paying the bills, leisure.

Of course you could choose to vegetate in front of a screen, but you're also free to walk the hills, create art, create machines, write software or poetry, or whatever.

> I personally would hate to spend all my time on leisure activities. I would still be working as hard but on my passion projects.

I would count that as leisure, why wouldn't that count?

Leisure isn't laying on a couch eating grapes; leisure is the freedom of deciding what to do with your time,.

In this context, "leisure" is used to mean the stuff you do outside of work and chores essentially. If you want to work on a side project, that would fall under leisure time.

> I personally would hate to spend all my time on leisure activities. I would still be working as hard but on my passion projects.

So … you would spend all your time on your chosen leisure activities—your passion projects. That is the point.

I second this. Leisure is not all it is cracked up to be.

> But as we lower that bar, the value a computer programmer has may begin to diminish.

As a developer, I feel like it's the opposite. If anything, it's going to raise the bar by replacing low skilled coders who are only good at cheaply producing piles of future tech debt, GPT4 will do that sort of work for 'free'. On the other hand it's going to empower non-programmers to write or debug low complexity scripts on their own, which is great.

I don't believe that any sort of "AI" short of AGI is going to replace developers doing "real work" on even medium complexity codebases. The reason for that is simple - GPT lacks logical reasoning capabilities. It can 'fake' them on the surface level, it might even fake them really well for extremely common problems, but as soon as you prod deeper or tack on an extra requirement or two, it starts spinning in circles indefinitely.

"Real" long-term software development is about taking (often poorly defined) requirements written in natural language, evaluating them in the scope of an existing software system, finding solutions that preserve behavior of the system (without introducing new bugs), and lastly implementing those changes in a way that follows the boundaries and abstractions defined in your system.

A single feature often touches many files at once. During this process you might spot a chance to introduce new abstractions or remove unnecessary ones. Unless you're planning on completely replacing developers with AI, the codebase must remain maintainable by mere mortals after you complete the next 50 revisions using AI that's mindlessly hacking away at your current codebase.

All of this requires logical reasoning capabilities that are, in my opinion, many orders of magnitude outside of GPT4's reach and you need those to maintain a software system for many years - decades even. That's before we get into the day to day communication and coordination effort required to sync with stakeholders, clarify requirements, and so on.

If I had to compare it to the security field, I believe that GPT will replace developers as much as automated vulnerability scanning tools have replaced pentesters. ;)

You used to have to be an expert/specialist to have direct access to a computer, or understand precisely how to give them commands to work with them over an expensive connection (and on a limited basis in terms of time and how many could do it).

That scarcity wasn't a good thing. People are drastically more empowered with computing today, the barriers were dropped in a rather extreme way (compared to where it started), and that is a good thing.

These are conceptual groupings of advancement. One layer of complexity is going to be removed (broadening access), there will be further advancement and complexity where the skilled group can navigate to in order to best unleash what they're capable of.

Writing the vast majority of code isn't special or super difficult (it's time consuming and mildly difficult at worst; quite obviously there are exceptions to that). There are a million plus software developers in the US alone. It's time to add new layers of complexity and advancement. Those million software developers will have to evolve, again, myself included.

I’m open to suggestions, at the end of the day it’s what works that matters, it’s just that every other system ever tried is minimum an order of magnitude worse. I’m not particularly a fan of the ‘red in tooth and claw’ brand of capitalism in the US. It’s a bit raw for me, but undeniably very dynamic.

Meanwhile in my lifetime capitalism has raised hundreds of millions of people out of poverty and into solidly middle class lifestyles globally.

My wife is Chinese so I’ve seen first hand how economic liberalisation has transformed that country. The Chinese brand of capitalism is severely flawed, largely due to an almost nonexistent rule of law, but compared to what they had before its night and day. It’s a shame the CCP seems to be thoroughly mismanaging it at the moment.

Proto capitalism is what lead up to feudalism. The Roman empire didn't have modern capitalism with banking, company shares and insolvency but what they did have is a silver/gold standard with debt bondage. Debt bondage is still a thing in Pakistan and the dynamic is exactly the same. You borrow a little money, the interest payments become overwhelming, you are forced to work on behalf of the creditor who can decide your wage and the interest rate.and whether they pay you in food and shelter or in money. There is no reason why they wouldn't abuse this power.

You see it all begins very benign. A farmer has a bad harvest. He needs to bridge over this year until the next harvest. So he borrows money with interest. When the next harvest fails then he becomes unable to pay and is forced to sell his land and perversively, hired to work on his own land. He has functionally become a serf. All it takes is for this to spread to the rest of society and you end up with feudalism.

In other words, feudalism is what happens when proto capitalism is exaggerated to an extreme extent.

If you were to exaggerate modern capitalism I am not sure what would happen. Maybe hyperinflation, maybe world war 3, maybe a reform of the banking system, maybe mass starvation or maybe a revolution or maybe modern feudalism except serfs aren't allowed to own intellectual property or shares in companies.

So no, I don't think the problem is that we don't have capitalism. We clearly have more capitalism than we want or need. Robber barons are an intrinsic part of capitalism.

The answer is a market economy without capitalism. I don't even see why anyone would care about capital obsession in a free market anyway. That sounds inherently autocratic. At some point in the capital accumulation process one person owns the entire economy, how does that not make them a monarch eventually?

I’d say this AI and all those that follow are children of capitalism.

It’s all about consolidating power and wealth as quickly as possible. Now we’re in an AI arms race. Sounds terrifying? It is.

America might tear itself apart if it doesn’t slow down a bit. Right now, there’s no one behind the wheel.

>The amount of misery under "capitalism" says there's got to be a better way.

I'm not quite sure why people think that a misery free world ought to be possible and/or the default.