An open letter to suspend work on AI

[Quentin111]

Pause Giant AI Experiments: An Open Letter.​

Bengio, Musk and Wozniak called for the suspension of training of AI systems more powerful than GPT-4.

 

[occupied]

They are not putting this genie back in the bottle.

 

[Anathema]

My first question is… how do we define “AI”? I work for an electric utility company. We’re working on implementation of a metering system (AMI) that would provide us with 24/7 real time usage info. We’re also bringing an automated FLISR (fault location, isolation, sensing and restoration) system online in the near future. Do these qualify as AI? FLISR is designed to “assist” (aka replace) Distribution Control Operators. AMI could replace most of our meter readers.

 

[Godboy]

My first question is… how do we define “AI”? I work for an electric utility company. We’re working on implementation of a metering system (AMI) that would provide us with 24/7 real time usage info. We’re also bringing an automated FLISR (fault location, isolation, sensing and restoration) system online in the near future. Do these qualify as AI? FLISR is designed to “assist” (aka replace) Distribution Control Operators. AMI could replace most of our meter readers.

AI can solve a million different problems in a sophisticated and nuanced way. Its an expert doctor, engineer, mathematician, writer, case researcher, etc all at the same time. Go to Chat GPT and play around with it. Its an expert in every niche subject you are into. Are you into the "Hope Diamond"? Ask Chat GPT. Hell, ask it to write you a story about Bruce Lee stealing it. You will be shocked by how smart it is.

 

[mak2]

My first question is… how do we define “AI”? I work for an electric utility company. We’re working on implementation of a metering system (AMI) that would provide us with 24/7 real time usage info. We’re also bringing an automated FLISR (fault location, isolation, sensing and restoration) system online in the near future. Do these qualify as AI? FLISR is designed to “assist” (aka replace) Distribution Control Operators. AMI could replace most of our meter readers.

I do not understand either. We train for ACLS on a website that does thousands of algorithms depending on what you do. It is surprisingly effective and no one dies when you practice. It was discussed as AI back when it first came out. But it is not, it just is programmed if you do this (like give a drug) your patient does that (like increased HR) but it is programmed, The program does not have independent judgement and/or wisdom.

 

[Anathema]

AI can solve a million different problems in a sophisticated and nuanced way. Its an expert doctor, engineer, mathematician, writer, case researcher, etc all at the same time. Go to Chat GPT and play around with it. Its an expert in every niche subject you are into

I tend to avoid technology which is intended to “out/think” me. There’s no Alexa in my home. I don’t use Siri on my phone. There will never be an AMI meter on my house or any “smart house” technology in it. I can think for myself and make my own decisions, thank you very much

 

[Godboy]

I tend to avoid technology which is intended to “out/think” me.

All technology outthinks us. Certainly when it comes to calculations and such.

 

[occupied]

There is great risk in AI becoming a black box technology that no one really understands but that is exactly where it is headed.

 

[SeaMajor7]

GIGO forever.

 

[Moonglow]

This is just like The Terminator all over again!

 

[shockedcanadian]

It's similar to nuclear warheads. If society will not/could not force the world to rid themselves of missiles that can melt all of humanity, A.I development will not stop. This is especially true since Americas adversaries wll not stop either so it is a competition now for who can replace humanity first. We need to rethink capitalism IMO as this development matures.

 

[Moonglow]

Can AI be meglomaniacs?

 

[evenflow1969]

Pause Giant AI Experiments: An Open Letter.​

Bengio, Musk and Wozniak called for the suspension of training of AI systems more powerful than GPT-4.

It is a bigger threat to humanity than nukes. They will not stop developing though. They will be to afraid someone will beat them to it.

 

[Anathema]

All technology outthinks us. Certainly when it comes to calculations and such.

I tend to separate mathematical calculations from decision making and critical thinking skills.

Just because 68 degrees may be the “optimal” thermostat temperature doesn’t mean that’s what I want it set to. If I’m willing to be a little colder I’ll set it to 65. If I’m willing to pay the costs to keep it at 70, that's up to me.

 

[scruffy]

AI can solve a million different problems in a sophisticated and nuanced way. Its an expert doctor, engineer, mathematician, writer, case researcher, etc all at the same time. Go to Chat GPT and play around with it. Its an expert in every niche subject you are into. Are you into the "Hope Diamond"? Ask Chat GPT. Hell, ask it to write you a story about Bruce Lee stealing it. You will be shocked by how smart it is.

It lies too. Probably built its database from Wiki pages.

 

[Quentin111]

Bill Gates says calls to pause AI won't 'solve challenges'.​

 

[Quentin111]

It is a bigger threat to humanity than nukes. They will not stop developing though. They will be to afraid someone will beat them to it.

In principle, for the sake of saving all of humanity, I could start working on explainable artificial intelligence (XAI) even tomorrow. Only for this I need a good mathematical education, and to get it I need about two million dollars (for example, studying at Harvard costs just about a million dollars). A little about myself: a graduate of the Ufa school with a specialization in physics and mathematics No. 114, graduated from the university with honors, with a red diploma without fours. During his studies, he was a laureate of a nominal scholarship named after the President of the Russian Federation B. N. Yeltsin. I have been interested in and engaged in artificial intelligence since 1991.

 

[mak2]

This is silly scary stuff. I truly do not believe a computer is intelligent, it is programmed. And I don't pretend to understand what AI really is, but we are not there yet. Don't be scared.

 

[evenflow1969]

In principle, for the sake of saving all of humanity, I could start working on explainable artificial intelligence (XAI) even tomorrow. Only for this I need a good mathematical education, and to get it I need about two million dollars (for example, studying at Harvard costs just about a million dollars). A little about myself: a graduate of the Ufa school with a specialization in physics and mathematics No. 114, graduated from the university with honors, with a red diploma without fours. During his studies, he was a laureate of a nominal scholarship named after the President of the Russian Federation B. N. Yeltsin. I have been interested in and engaged in artificial intelligence since 1991.

So right about the time they were playing with fuzzy logic. Is when you entered. As far as getting that education and fighting against AI I think you are to far behind. By the time you know what you need to know computers will be gaining intelligence at a geometric rate if they are not now.

 

[Quentin111]

So right about the time they were playing with fuzzy logic. Is when you entered. As far as getting that education and fighting against AI I think you are to far behind. By the time you know what you need to know computers will be gaining intelligence at a geometric rate if they are not now.

Here is a good weighted assessment based on the accumulated opinion of many experts on this issue. The confidence interval lies in the range from 2026 to 2041. The median is 2031. So if I get a master's degree in mathematics in 8 years, then there is a chance that I will have time to save the world.