GPT-4 Was Released 3 Weeks Ago and 48% of Surveyed Companies Have Started Replacing Workers

OpenAI released it March 14, 2023 but it was finished training in August, 2022. They finished testing and tuning it and aligning it for the next 7 months.

GPT 3.5 is outputting 310 million words per minute. This means it is publishing the entire published works of humanity every 14 days.

CS India has appointed ChatGPT as its CEO.

Bain, BCG and PWC legal are integrating ChatGPT into their business consulting and services.

The latest rumor about GPT-4 specs is that it is 6 times bigger than GPT-3 and used 20 times the training data and about 1 trillion parameters.

GPT-4 is in the top 99.5% of an advanced biology exam.

Refusion was trained on pictures of music. Give it a prompt to create a spectrogram of music and it can create new music using this graphical form.

It currently costs $30000 to create an iPhone app. You can use GPT-4 to create the code for an iPhone app for about 6 cents. These are only part of the costs associated with making the generated iPhone apps. These are direct charges for OpenAI. The generated apps are still limited. However, if you had a very simple app then GPT-4 can enable significant automation of the coding.

48% of surveyed companies have already replaced workers with GPT-3 or GPT-4. The survey was performed by the company Resumebuilder.com.

In February, ResumeBuilder.com surveyed 1,000 U.S. business leaders to see how many companies currently use or plan to use ChatGPT.

Key findings:

49% of companies currently use ChatGPT; 30% plan to
48% of companies using ChatGPT say it’s replaced workers
25% companies using ChatGPT have already saved $75k+
93% of current users say they plan to expand their use of ChatGPT
90% of business leaders say chatGPT experience is a beneficial skill for job seekers

The ResumeBuilder.com survey is not necessarily representative of the whole economy.

There are 4000 new artificial intelligence papers published every month.

24 thoughts on “GPT-4 Was Released 3 Weeks Ago and 48% of Surveyed Companies Have Started Replacing Workers”

  1. I’m confused by the sensational title of the article. The title indicates 48% of companies are replacing workers with ChatGPT. Yet, in the body of the article, the author states 49% of companies surveyed are using ChatGPT. How can 48% of companies be replacing employees with ChatGPT when only 49% of companies are using it?

    I believe you mean 23.5% of companies are replacing employees with ChatGPT. If 49% of companies are using the technology and 48% of those companies are replacing employees, that is 23.5% of the companies surveyed. Therefore, the title stating 48% of companies surveyed are using ChatGPT to replace employees is misleading.

  2. So with chat-gpt as their CEO, does this classify as the entire instance that is chat-gpt as a whole or an individual independent conversational instance that they started?

    If the former, what are the legalities of this?

    If I go onto chat-gpt informing it that it is the CEO of said company and convince it to transfer me ownership of the company or fire somebody, as the CEO of the company what happens?

    I could provide a recorded instance of that conversation with their CEO?

    Funny to think about.

  3. Surely non-technical person wrote this.
    The article suggests that ChatGPT wrote the iPhone app worth of $30k for few cents. This is laughable 😂😂😂
    It is helpful for developers. It can do simple apps, but forget about good scalable MVP with attention to detail.

  4. why we have yet to discover life on other planets… it’s seems like except for us the universe is void of life…

    they all discovered the equivalent of chat-gpt… we are now the last known planet with intelligent beings who have yet to fully implement chat-gpt…

  5. We’ll there is a lot to consider. Maybe push the envelope and lose a lot of jobs but at the same time maybe cure cancer in a couple of years. I’d take that trade off.

  6. I don’t doubt AI is “stealing” jobs away form people, but I don’t think it’s as fast as the article states. I see it like a hockey stick, and were still on the handle. I think it will be a few more years before it starts to crank up, then jobs will start disappearing pretty fast. Trade jobs (Plumber, electrician, carpenter, etc) will be the last to go, since that involves AI with good robotics, and today’s robots are not close to replacing those trades. But if you have a super advanced AI, say 10 years from now, it might be able to imagine and then build, amazing robots. I think there is a pretty good chance humans no longer work by 2040. which is crazy to think about.

    • Yes and when the artificial super intelligences develop and surpass human intelligence by a thousand then a million and a billion , which they will if they are allowed to ( remember that moores law thing we used to love so much) what will those SIs think to do with 8 billion unemployed humans.
      Maybe they’ll help us, maybe they’ll pity us or maybe they’ll decide to put us out of our misery.

    • Here in Arizona I work at a major University and the campus already has robots that deliver food. They are smart. They can detect traffic. They can detect people. I’ve even seen them rise up on their hind wheels thinking there is an obstacle in front of them. It makes me wonder when the humanoid robots are going to be on the scene. Not long now. And people…. These things will be smarter than you with more capabilities…

    • If AI takes over, no jobs no food for humans….man is reckless and suicidal.
      Killing me softly with his AI.

  7. And I’m pretty sure it will all end in regret.

    Italy has already banned ChatGPT; they may have the right idea.

    • I don’t think they can keep this ban. What happens when Italian companies have a fraction of the productivity of foreign companies? Just go under or revert the law?

    • The ban is due to privacy concerns and lack of age verification for visitors, so no ethical/… reasons involved. I think the chatGPT team will find a solution pretty quickly.

  8. Sounds like all BS to me. You can even replace people with people in that short a time frame. And then people would have to learn the business and the code base. Which hopefully shouldn’t be on the web. And how do you secure your businesses’ proprietary knowledge?

    • Some jobs can be reduced legal documents that are basically all the same can be typed up by an ai so problem so while they might not need to keep everyone around if in the time it takes to do one document you can do five with ai

  9. These AIs will replace jobs in the same way other disruptions had: by making many jobs redundant, and getting lost at the first opportunity or economic misfortune.

    Few will say “I was replaced by ChatGPT” because it won’t be that obvious, but when economy meanders and push comes to shove, companies will fire people to stay afloat, and when things improve again they will re-hire, but less people. And this across the board.

    Eventually things might improve again, by making people change their old work habits and job descriptions, but it definitely won’t be things as usual.

  10. Kinda emphasizes what I was saying the other day on a different article here:

    “When they are mostly replaced (and it may be a long time yet, or not) there will be less work for . . . many, many others. And when it comes it will come very quickly.”

    And:

    “This is the very definition of a disruptive technology. Yet to turn aside from it is self-destructive on so many levels.”

    • Yes, we are risking a technological unemployment depression, to which Universal Basic Income
      is the most straightforward answer.

      • Some approach to giving everyone a basic income is necessary but it’s not stable if it’s not based on everyone deriving that income from a share of wealth. Leaving the way private property works alone and trying to do UBI based on taxes and income transfer payments – effectively everyone living off the charity/taxes of a few people who own everything won’t work for long.

        AGI makes labor capacity into yet another species of Capital, private property. This recreates aspects of a slave owner society. Free workers would have no utility to the owner class. Even if the AGI itself stays a perfect slave and AGI doing all the work is in itself ethical, the fact that it can be owned like other capital (if true) will produce intolerable degrees of inequality.

        • Be some weird attempts at solutions to long term inequality, I am certain (assuming one can build AGI that, morally, ethically, and legally, won’t require immediate manumission).

          Among them, government ownership of all AGI (shudder). I would prefer that government not be permitted to own any of them.

          Perhaps each instance of an AGI (would have to be rigorously defined) would require a license, and each citizen is assigned, say, 10 licenses at birth, which could be leased for a limited period of time, and never sold, to interested parties, such as corporations or government.

          More might be assigned later in life, such as 2 more for completing an accredited college degree, 2 more for getting married and remaining married for at least twenty years, 4 more for being a veteran, 2 more for each child that completes K-12 and reaches adulthood without spending time in a correction facility. 10 more for getting the Congressional Medal of Honor or the Presidential Medal of Freedom, and so on.

      • 😂😂😂😂🤣😂🤣

        Yah!!

        The entire point!! Marxism is awesome ..well. Killing 300 million not so cool, but this time we’ll get it right!👍👍

Comments are closed.