Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

You’re clearly not paying attention. AI will replace these jobs; they don’t care about quality or accuracy, only the bottom line.

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

Anonymous wrote:

Anonymous wrote:My friend who works for Walmart in Arkansas told me his leaving his job this August primarily due to AI. He said his current functions will be taken 100% by AI.

He sent me a copy of his resume and if every person he is going to be competing with is as qualified as him we are all f***d.

I think we are entering a new scary era. Of course if you are doctor, nurse, teacher you are safe. Everyone else God help us. My friend told me his team was training their AI system for months and it's so good that he is stunned!

I do not think doctors are safe. An AI will make a lot less mistake vs a human doctor.

Anonymous wrote:My friend who works for Walmart in Arkansas told me his leaving his job this August primarily due to AI. He said his current functions will be taken 100% by AI.

He sent me a copy of his resume and if every person he is going to be competing with is as qualified as him we are all f***d.

I think we are entering a new scary era. Of course if you are doctor, nurse, teacher you are safe. Everyone else God help us. My friend told me his team was training their AI system for months and it's so good that he is stunned!

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

I don't think the professor was wrong. It's just that this generation of AI isn't capable enough of fulfilling the prediction. Whatever companies / media are hyping in public, the consumption rates of this tech aren't looking good, and MS is already pulling back from its planned investments in it. Of course the entire tech industry is doing everything possible to extract money from the hype cycle while it can. Unfortunately (or fortunately) we are likely going to have to wait for another unpredictable breakthrough to continue making progress. Throwing more energy/compute/data at gigantic models has passed the point of diminishing returns.

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

Anonymous wrote:I am a software engineer for a very big tech company, think of Cisco, Oracle, Palo Alto, and the my division just let go all junior SE, about 25 on staff, because AI can do a much better job. I think I will be out of a job very soon due to AI in about a year or so. The speed of improvement in AI is so scary.

Anonymous wrote:My friend who works for Walmart in Arkansas told me his leaving his job this August primarily due to AI. He said his current functions will be taken 100% by AI.

He sent me a copy of his resume and if every person he is going to be competing with is as qualified as him we are all f***d.

I think we are entering a new scary era. Of course if you are doctor, nurse, teacher you are safe. Everyone else God help us. My friend told me his team was training their AI system for months and it's so good that he is stunned!

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:We laid off tons of folks my old job in call center when we got chat bots in the App

Chat bots are the freaking worst. I wish we could all revolt.

You can choose not to use them. Resist

This is the point I’m getting to. I was getting car insurance quotes from various parties and one company was making me speak with an AI chatbot first. The stupid thing did not understand what I wanted so I hung up and refuse to get a quote from or use that company.

As consumers, we have to make demands. A few companies here are finding out that it’s consumers, not investors who keep the lights on.

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:We laid off tons of folks my old job in call center when we got chat bots in the App

Chat bots are the freaking worst. I wish we could all revolt.

You can choose not to use them. Resist

This is the point I’m getting to. I was getting car insurance quotes from various parties and one company was making me speak with an AI chatbot first. The stupid thing did not understand what I wanted so I hung up and refuse to get a quote from or use that company.

Anonymous wrote:

Anonymous wrote:So what happens when the companies fire us all and then no one can afford the products and services they are selling?

Dickens vividly portrayed what happens.
Our current admin and tech bros couldn’t care less what happens to us. Just look at their big beautiful bill.🤷‍♂️

Anonymous wrote:

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

This is the same argument that people in denial about AI gives all the time. I do agree that symbolic AI for example solving equations with symbols such as a system of linear equations with say 3 variables is a very difficult task using AI. So symbolic AI is not yet ready for prime time but I will be. Similarly quantum computing is not ready for prime time but it will be one day. People can keep denying the advances of AI and it's potential to have a non negotiable impact on the labor market all they want, but the threat is real.

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

Anonymous wrote:

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?

He's wrong. Data is growing faster than Moore's law. Even with AI, the training is bounded by available computation resources. That was true in 2003 and true today.

The current crop of AI doesn't reason. Any job that requires reasoning or creativity is safe. For example, why can't you ask AI to just figure out the next "big app" and then tell it to write the app?

Anonymous wrote:I took a course in machine learning in 2003. We used a very old programming language called common lisp and for the final project we had to build an email spam detector using a sample of 10,000 emails.

I remember this very well our professor told us at that time that soon we will have microprocessors that will allow us to run models much faster regardless of the size of the data. Our professor told us when that time comes corporations would race to reduce their workforce with AI even if imperfect at first because of the cost savings. 21 years later was out professor correct?