ChatGPT is an Open Source AI
- Thread starter ZNP
- Start date
-
Christian Chat is a moderated online Christian community allowing Christians around the world to fellowship with each other in real time chat via webcam, voice, and text, with the Christian Chat app. You can also start or participate in a Bible-based discussion here in the Christian Chat Forums, where members can also share with each other their own videos, pictures, or favorite Christian music.
If you are a Christian and need encouragement and fellowship, we're here for you! If you are not a Christian but interested in knowing more about Jesus our Lord, you're also welcome! Want to know what the Bible says, and how you can apply it to your life? Join us!
To make new Christian friends now around the world, click here to join Christian Chat.
6 Do not be anxious about anything, but in every situation, by prayer and petition, with thanksgiving, present your requests to God.
7 And the peace of God, which transcends all understanding, will guard your hearts and your minds in Christ Jesus.
8 Finally, brothers and sisters, whatever is true, whatever is noble, whatever is right, whatever is pure, whatever is lovely, whatever is admirable—if anything is excellent or praiseworthy—think about such things.
9 Whatever you have learned or received or heard from me, or seen in me—put it into practice. And the God of peace will be with you.
7 And the peace of God, which transcends all understanding, will guard your hearts and your minds in Christ Jesus.
8 Finally, brothers and sisters, whatever is true, whatever is noble, whatever is right, whatever is pure, whatever is lovely, whatever is admirable—if anything is excellent or praiseworthy—think about such things.
9 Whatever you have learned or received or heard from me, or seen in me—put it into practice. And the God of peace will be with you.
AI Experts Revise Predictions by 48 YEARS - We're in the endgame now!
Two major milestones
Looking at full automation of labor and HLMI -- High level machine intelligence.
Better, faster, cheaper and safer -- aggregate forecast by 2047! Down 13 years of predictions previous to 2022.
Compare this with the PC revolution. In 1981 the IBM PC was invented. By the year 2000 half the homes in America had a computer. Facebook and Youtube came out in 2004/2005. In 2011 Watson won Jeopardy, the start of the AI age. In 2016 AlphaGo beat Lee Sedol, the beginning of AI teaching itself.
It took 19 years for computers to get into half the homes in the US. Therefore I think it is very reasonable (due to Moore's law) that AI will impact 50% of jobs by 2030. The problem with human predictions is that they don't know everything that is taking place and you have to in order to make an accurate prediction.
So then what does 50% mean? It probably means no new hires and people in the industry only keep their jobs if they are adept at using AI. This is very bad news for anyone who is a freshman in high school or younger. In fact I would think the next seven years will be a very hard time for people trying to get an entry level job, probably impossible in one to two years. That is very bad news for anyone who is a sophomore in college or younger right now.
Two major milestones
Looking at full automation of labor and HLMI -- High level machine intelligence.
Better, faster, cheaper and safer -- aggregate forecast by 2047! Down 13 years of predictions previous to 2022.
Compare this with the PC revolution. In 1981 the IBM PC was invented. By the year 2000 half the homes in America had a computer. Facebook and Youtube came out in 2004/2005. In 2011 Watson won Jeopardy, the start of the AI age. In 2016 AlphaGo beat Lee Sedol, the beginning of AI teaching itself.
It took 19 years for computers to get into half the homes in the US. Therefore I think it is very reasonable (due to Moore's law) that AI will impact 50% of jobs by 2030. The problem with human predictions is that they don't know everything that is taking place and you have to in order to make an accurate prediction.
So then what does 50% mean? It probably means no new hires and people in the industry only keep their jobs if they are adept at using AI. This is very bad news for anyone who is a freshman in high school or younger. In fact I would think the next seven years will be a very hard time for people trying to get an entry level job, probably impossible in one to two years. That is very bad news for anyone who is a sophomore in college or younger right now.
I want to point out that AlphaGo was at the 85% level of Go players when it began teaching itself and in a matter of 3 months was better than the world's best AlphaGo player. Since that time computer power has increased 4 fold. Perhaps today it would take AlphaGo 1 month to make that improvement.
Now, imagine a Fortune 500 company using AI for accounting hooked up to their powerful mainframe supercomputer, feeding their proprietary software with all the updated tax laws and making the program focused on doing this one thing very well. If a Fortune 500 company has 1,000 accountants surely this AI could replace 200 the first year with all final sign offs being done by people. By the second year you could cut your work force in half, and by the third you are down to 20%, and then 10%. This is achieved by hiring freeze, early retirement and then layoffs.
The same thing can be done with programmers and coders at google and other tech companies.
It can be done with the legal department.
Small companies don't have the resources or financial incentive to move this quickly (eliminating 200 jobs making 100k a year saves them 20 million a year), however, they will have to compete with these fortune 500 companies and they will lose. Also, these fortune 500 companies can sell subscription to their mainframes for the legal AI, the CPA AI, the programming AI, etc.
Now imagine what happens when you put these programs on a Quantum computer. How much faster, and smarter will it be?
Now, imagine a Fortune 500 company using AI for accounting hooked up to their powerful mainframe supercomputer, feeding their proprietary software with all the updated tax laws and making the program focused on doing this one thing very well. If a Fortune 500 company has 1,000 accountants surely this AI could replace 200 the first year with all final sign offs being done by people. By the second year you could cut your work force in half, and by the third you are down to 20%, and then 10%. This is achieved by hiring freeze, early retirement and then layoffs.
The same thing can be done with programmers and coders at google and other tech companies.
It can be done with the legal department.
Small companies don't have the resources or financial incentive to move this quickly (eliminating 200 jobs making 100k a year saves them 20 million a year), however, they will have to compete with these fortune 500 companies and they will lose. Also, these fortune 500 companies can sell subscription to their mainframes for the legal AI, the CPA AI, the programming AI, etc.
Now imagine what happens when you put these programs on a Quantum computer. How much faster, and smarter will it be?
People are concerned about AGI -- artificial General intelligence.
Well this breakthrough is ATI -- Artificial training intelligence. The AI can watch someone else do it and then learn simply by watching. Imagine how many training videos on Youtube there are! Imagine you have a mechanic shop and you show the robot how to change oil, then show it how to change a tire, and then show it how to fix a broken light, etc. You can imagine Midas having robots that fix transmissions, and AAMCO and all the other transmission chain companies.
Imagine these chains that change your oil and other fluids. Take one day to train the AI letting it watch twenty videos. One owner could have two or three bays with two or three robots. If some car it has never worked on comes in the owner shows the AI how to do it.
These robots cost around 120k, about the cost of two or even three employees, but the robot can work four shifts a week, so it can easily justify the cost in one year. That is 100% ROI, generally anything over 25% is considered a great investment.
Obviously you could train it to work in the kitchen of a McDonalds or some other fast food restaurant. If you are open 24 hours a day, 7 days a week you would have a much more reliable employee, never late.
Well this breakthrough is ATI -- Artificial training intelligence. The AI can watch someone else do it and then learn simply by watching. Imagine how many training videos on Youtube there are! Imagine you have a mechanic shop and you show the robot how to change oil, then show it how to change a tire, and then show it how to fix a broken light, etc. You can imagine Midas having robots that fix transmissions, and AAMCO and all the other transmission chain companies.
Imagine these chains that change your oil and other fluids. Take one day to train the AI letting it watch twenty videos. One owner could have two or three bays with two or three robots. If some car it has never worked on comes in the owner shows the AI how to do it.
These robots cost around 120k, about the cost of two or even three employees, but the robot can work four shifts a week, so it can easily justify the cost in one year. That is 100% ROI, generally anything over 25% is considered a great investment.
Obviously you could train it to work in the kitchen of a McDonalds or some other fast food restaurant. If you are open 24 hours a day, 7 days a week you would have a much more reliable employee, never late.
If I am aware of this innovation you can be sure the higher ups at McDonalds and Walmart and Midas are. They all have the resources to buy and train these robots and once they have one trained they can buy a whole bunch of the robots pretrained. Since all these stores are pretty much identical it would be very easy to get these robots working in every Walmart or McDonald's or Midas, etc. I assume these companies have training videos and the could probably make an entirely new set of training videos with the robot doing the task that could be part of the AI's data set. I would think it would take one month for one of these corporations to train the robots. Then a large order might get them the 10% discount, so instead of $120k maybe they get the robot for $108k.
You begin with the stores that can use these robots 24 hours a day, seven days a week. After 1 year you can sell the used ones to stores that are opened fewer hours for half price. 2024 will see an explosion of the use of AI.
I thought we might still have two more years before these jobs go, but now I think this is the year they will be taken.
You might not want to replace your workers with robots but you won't have any choice. Walmart's biggest expense is labor, this could cut that cost by 75% at least, if not more. No one will be able to compete with them unless they too switch to robots.
You begin with the stores that can use these robots 24 hours a day, seven days a week. After 1 year you can sell the used ones to stores that are opened fewer hours for half price. 2024 will see an explosion of the use of AI.
I thought we might still have two more years before these jobs go, but now I think this is the year they will be taken.
You might not want to replace your workers with robots but you won't have any choice. Walmart's biggest expense is labor, this could cut that cost by 75% at least, if not more. No one will be able to compete with them unless they too switch to robots.
They don't all cost that much. And the prices will fall quickly as more are purchased. But technology had been replacing jobs and creating new ones since the garden. God Himself put plant based clothing stores out of business almost as soon as they started. It was always God's plan to move from the garden into a city.
Bill Gates Shocking New 2024 AI Prediction ( AGI + AI Agents)
"We are at the beginning of this transition right now"
"AI is about to supercharge the innovation pipeline" -- This is what I think we need to be fully alert because the rate of change will increase exponentially. It can speed up the process of research and of development.
"18-24 months away from significant use of AI in the US" -- This is what I have been saying. In six months the Fortune 500 companies can all have a plan on how they will adopt AI and most likely this will include a hiring freeze as they anticipate this dramatic increase in use. Why train a new hire when instead you should be training your AI. I suspect in boardrooms all over the country that is what is going on. Hiring freeze, have a plan to adopt and train AI and in so doing allow the workers you have now to do 5% more. 12 months from now they will probably institute a wage freeze along with incentives for early retirement. 18 months from now rapid increase in the adoption of AI, leaving 20% of your workforce with nothing to do. By 24 months there will be many, many pink slips going out reducing the workforce on many companies by 40%. Compare that with the Great Depression when 25% of Americans were unemployed. It will be worse than what we are seeing in China right now. This is why all the countries are pushing for war, they know their population could be reduced by 50% over the next two years and that would actually eliminate problems.
Over the next few years Siri will be on steroids. Most people will lose the ability to do anything, simply tell Siri what to do and she'll do it. Schedule something, plan an itinerary, map your route, buy tickets, reserve a seat at the restaurant, etc. There will be a new generation that has no idea how to do anything on the computer, they simply tell Siri. Agents will be like a really good butler. ChatGPT is getting 10 million requests per day. This would be like a Butler who got 456 requests per day for sixty years. This is what the training is for ChatGPT, it is like the amount of training 100 butlers got over their entire career every day.
This will have a major impact in healthcare. Think about all the trips you made to the doctor in your life, how often did it involve surgery and how often was it simply talking to the doctor, answering questions, having some tests prescribed? You will probably still need a person to take blood, but instead of waiting for thirty minutes in the waiting room you would sit before the AI, it would take blood pressure, measure your weight, and discuss your health, and prescribe various tests. In fact you could access this before you even go to the doctor and get a print out from the AI that you show them when you check in.
"We are at the beginning of this transition right now"
"AI is about to supercharge the innovation pipeline" -- This is what I think we need to be fully alert because the rate of change will increase exponentially. It can speed up the process of research and of development.
"18-24 months away from significant use of AI in the US" -- This is what I have been saying. In six months the Fortune 500 companies can all have a plan on how they will adopt AI and most likely this will include a hiring freeze as they anticipate this dramatic increase in use. Why train a new hire when instead you should be training your AI. I suspect in boardrooms all over the country that is what is going on. Hiring freeze, have a plan to adopt and train AI and in so doing allow the workers you have now to do 5% more. 12 months from now they will probably institute a wage freeze along with incentives for early retirement. 18 months from now rapid increase in the adoption of AI, leaving 20% of your workforce with nothing to do. By 24 months there will be many, many pink slips going out reducing the workforce on many companies by 40%. Compare that with the Great Depression when 25% of Americans were unemployed. It will be worse than what we are seeing in China right now. This is why all the countries are pushing for war, they know their population could be reduced by 50% over the next two years and that would actually eliminate problems.
Over the next few years Siri will be on steroids. Most people will lose the ability to do anything, simply tell Siri what to do and she'll do it. Schedule something, plan an itinerary, map your route, buy tickets, reserve a seat at the restaurant, etc. There will be a new generation that has no idea how to do anything on the computer, they simply tell Siri. Agents will be like a really good butler. ChatGPT is getting 10 million requests per day. This would be like a Butler who got 456 requests per day for sixty years. This is what the training is for ChatGPT, it is like the amount of training 100 butlers got over their entire career every day.
This will have a major impact in healthcare. Think about all the trips you made to the doctor in your life, how often did it involve surgery and how often was it simply talking to the doctor, answering questions, having some tests prescribed? You will probably still need a person to take blood, but instead of waiting for thirty minutes in the waiting room you would sit before the AI, it would take blood pressure, measure your weight, and discuss your health, and prescribe various tests. In fact you could access this before you even go to the doctor and get a print out from the AI that you show them when you check in.
Imagine replacing the doctor that diagnoses your issue with the technician that draws blood. AI will replace the job that earns 200k a year while leaving the job that is paid 40k a year. It can do the job of the receptionist, and the guy who pulls your file, and the doctor who diagnosis you, and the person who checks your blood pressure and weight as well as the janitor. However, you will still have the technician drawing blood. So although it may only take 80% of the jobs it will reduce labor costs by 90%.
The doctor who diagnosis a patient doesn't have a valuable skill?
Truck drivers don't have valuable skills?
The mechanic who diagnosis the problem with the car doesn't have a valuable skill?
The lawyer who writes up a legal contract or will doesn't have a valuable skill?
The CPA who does your taxes doesn't have a valuable skill?
The fighter pilot who flies F16s doesn't have a valuable skill?
The Ranger sniper doesn't have a valuable skill?
Computer programmers don't have valuable skills?
Truck drivers don't have valuable skills?
The mechanic who diagnosis the problem with the car doesn't have a valuable skill?
The lawyer who writes up a legal contract or will doesn't have a valuable skill?
The CPA who does your taxes doesn't have a valuable skill?
The fighter pilot who flies F16s doesn't have a valuable skill?
The Ranger sniper doesn't have a valuable skill?
Computer programmers don't have valuable skills?
Technology has allowed us to bypass cognition.
As a teacher I saw this, kids had absolutely no clue how to do simple math, they plug the numbers into a calculator and write down whatever it says without any thinking or common sense. To illustrate this I was teaching a HS science class, and the metric system. Students had to measure their desk and write down the answer. One girl told me her desk was 100 meters long. When kids use calculators you get answers like that all the time.
Bookkeepers used to have to do calculations all day long, then they had to use calculators all day long. Now they simply are data entry clerks and the spreadsheets do the rest. However, with AI you can eliminate the data entry as well.
Then with google we saw kids would use google search to do all their homework. There are sites on the internet that do just that, they answer all the questions on various text books. We learned we had to take phones away from kids as they enter the building otherwise they would cheat all day long.
Think about the Regents exam. As a teacher I would have access to the answer key once the test began. All it would take is one corrupt teacher somewhere in the entire state to leak the answers on the internet. So imagine a student comes to the test 30 minutes late, we will still let them take it, but by that time they may have already gotten all the answers. These days anyone can photograph the answer key and post it in about 5 minutes. It seems to me with VPN and all these high tech scams this would be relatively simple thing to do, especially if hidden on some obscure website. Even if you locked up the answer key until the test was over any half decent teacher could do the exam in thirty minutes. Cheating scandals are a common thing in the SAT, and in high stakes exams.
Well now AI has taken this a whole new step further with a program that can do the research for you and write the paper for you. As a teacher I know that about the only way to keep the students from using this exclusively will be to homeschool. I also know that teachers will use it to grade the papers. If you can now bypass reading and writing and not just simple math, what will be left for human cognition? It used to be the way we dealt with this issue is make the assignment very specific so that there was no paper that could be plagiarized that met those specifics. At the very least the kid would have to paraphrase. But now with AI that will be. a snap. When I went to school we would get research papers assigned with two weeks or even a month to do them. Now, you get a paper, come home tell AI what you want and you have it in five minutes.
As a teacher I saw this, kids had absolutely no clue how to do simple math, they plug the numbers into a calculator and write down whatever it says without any thinking or common sense. To illustrate this I was teaching a HS science class, and the metric system. Students had to measure their desk and write down the answer. One girl told me her desk was 100 meters long. When kids use calculators you get answers like that all the time.
Bookkeepers used to have to do calculations all day long, then they had to use calculators all day long. Now they simply are data entry clerks and the spreadsheets do the rest. However, with AI you can eliminate the data entry as well.
Then with google we saw kids would use google search to do all their homework. There are sites on the internet that do just that, they answer all the questions on various text books. We learned we had to take phones away from kids as they enter the building otherwise they would cheat all day long.
Think about the Regents exam. As a teacher I would have access to the answer key once the test began. All it would take is one corrupt teacher somewhere in the entire state to leak the answers on the internet. So imagine a student comes to the test 30 minutes late, we will still let them take it, but by that time they may have already gotten all the answers. These days anyone can photograph the answer key and post it in about 5 minutes. It seems to me with VPN and all these high tech scams this would be relatively simple thing to do, especially if hidden on some obscure website. Even if you locked up the answer key until the test was over any half decent teacher could do the exam in thirty minutes. Cheating scandals are a common thing in the SAT, and in high stakes exams.
Well now AI has taken this a whole new step further with a program that can do the research for you and write the paper for you. As a teacher I know that about the only way to keep the students from using this exclusively will be to homeschool. I also know that teachers will use it to grade the papers. If you can now bypass reading and writing and not just simple math, what will be left for human cognition? It used to be the way we dealt with this issue is make the assignment very specific so that there was no paper that could be plagiarized that met those specifics. At the very least the kid would have to paraphrase. But now with AI that will be. a snap. When I went to school we would get research papers assigned with two weeks or even a month to do them. Now, you get a paper, come home tell AI what you want and you have it in five minutes.
Open AI RESPONDS To New York Times LAWSUIT
Consider AI and journalism. Many stories are simply reporting a press release. New park is opened, police arrest criminal, police seize drugs, somebody died, etc. Surely AI can do that.
Other stories involve investigation. I saw a movie called Bad Education about the Roselyn school district in Long Island with a massive case of embezzelment. Almost the entire investigation was based on looking at public records. Certainly AI could do that and probably could do ten times as many of those investigations.
Some journalists sit and listen to a press secretary like at the Whitehouse and then ask questions. Certainly AI can do that.
In addition to this journalists read their competitors and watch other news stories around the globe as well as subscribe to wire services. Certainly AI can do that.
Then of course you have others who will sit in front of a camera and read the stories, certainly AI using deep fake can do that.
Then of course you have editors who edit the copy given to them, certainly AI can do that.
My point is that although you would assume you can't replace all journalists, it does seem like replacing 80% of them shouldn't be too difficult.
Consider AI and journalism. Many stories are simply reporting a press release. New park is opened, police arrest criminal, police seize drugs, somebody died, etc. Surely AI can do that.
Other stories involve investigation. I saw a movie called Bad Education about the Roselyn school district in Long Island with a massive case of embezzelment. Almost the entire investigation was based on looking at public records. Certainly AI could do that and probably could do ten times as many of those investigations.
Some journalists sit and listen to a press secretary like at the Whitehouse and then ask questions. Certainly AI can do that.
In addition to this journalists read their competitors and watch other news stories around the globe as well as subscribe to wire services. Certainly AI can do that.
Then of course you have others who will sit in front of a camera and read the stories, certainly AI using deep fake can do that.
Then of course you have editors who edit the copy given to them, certainly AI can do that.
My point is that although you would assume you can't replace all journalists, it does seem like replacing 80% of them shouldn't be too difficult.