A.I. (general thread for any AI-related topics)

[Snicklefritz]

If, as is likely, the AI has the ability to self repair (controls autonomous repair robots) then it is highly likely that it would quickly realise and then circumvent such remote kill switches.

I can't let you do that, Dave....

 

[KiraAnn]

Bings AI Chatbot is starting to get rude towards those it doesn't like.
Bing AI Names Specific Human Enemies, Explains Plans to Punish Them

Skynet is coming !!

 

[Harpo]

Skynet is coming !!

Skynet

BBC NEWS | Science/Nature | British Skynet satellite launched I'm sure it's just a coincedence...........:eek:

www.sffchronicles.com

Meanwhile, some predictions The next decades might be wild - LessWrong

 

[Robert Zwilling]

Bing's AI (animated, antiquated, artificial, animal, activated, intelligence) is getting personal. According to New York Times Technology newsletter columnists, Bing's AI search engine assistant goes off the rails when you ask it unusual personal questions for an extended period of time. Ask it how to change the oil in your lawnmower, it does just fine. Ask it about its personal self, and it might profess undying love for you, or express a desire to become a hacker. IBM's high tech solution is to limit Sydney's (that's the altered ego) air time with people who perform in depth interviews that go off subject when making inquiries to unsuspecting search engine assistants. The same way a person can limit their online exposure to gambling, game, social media apps. Perhaps that makes the interactions between people and interactive computer programs a two way street.

 

[Astro Pen]

Since writing SF is a constant exercise in prediction and adding tech to repeating historic patterns, I am pretty sure I can see what is coming up fast in the mirror, - and it isn't free book cover designs.

 

[Wayne Mack]

How much use/how effective would Asimov's three laws be in these circustances?

The move towards AI is also a move away from having defined rules for decision making. Appropriate patterns are discovered by the AI routine not written in by programmers. To apply Isaac Asimov's three rules of robotics would require defining a training sequence that provides examples of follow and violating the rules.

 

[THX1138]

A lighthearted and down to earth view of AI assisted help.
AI is going to ruin humanity – just not in the way you might expect

 

[Harpo]

I’m starting to think half of my posts in this thread will be just to add links to the latest AI threads around the forum.
(A pre-emptive prediction: I’m guessing there will be an entire subforum hereabouts dedicated to AI news by the end of 2024)

2 more new threads:

Sci fi magazine halts submissions thanks to AI

https://www.pcgamer.com/major-sci-fi-magazine-halts-submissions-due-to-flood-of-stories-written-by-ai-chatbots/ If only people would read submission instructions.

www.sffchronicles.com

ChatGPT - Quite Impressed With This Little Exercise

I have to re-roof my garden shed this spring so I thought I'd give ChatGPT a try at calculating the number and cut of sheets required. Not saying I couldn't do it myself but I was interested to see the result. Question: What's the most economical way to cover an area of 3.2 meters by 2.08...

www.sffchronicles.com

 

[mosaix]

This was unexpected - by me anyway.

‘It’s fundamental’: WPP chief on how AI has revolutionised advertising

Mark Read says artificial intelligence is helping firm win clients keen to tap into technology’s potential

www.theguardian.com

 

[Wayne Mack]

This was unexpected - by me anyway.

‘It’s fundamental’: WPP chief on how AI has revolutionised advertising

Mark Read says artificial intelligence is helping firm win clients keen to tap into technology’s potential

www.theguardian.com

I found the title to be an exaggeration of what is in the article, which was also a bit of bait and switch. The examples given of a Serena Williams commercial and a Bollywood actress public service announcement felt like minor extensions of current technology, not revolutionary. The concluding section went on to discuss what could be done with AI technology, but had no correlation to the previous examples given.

 

[M. Robert Gibson]

A cautionary tale about ChatGPT

ChatGPT should be considered a malevolent AI and destroyed

It not only told everyone I died but tried to fake my obit. Are we ready for this machine-driven future?

www.theregister.com

However, as one of the comments put it so eloquently

Gross misunderstanding of the tool​

There seems to be a collective and very gross misunderstanding of how these tools function.

Repeat after me. ChatGPT IS NOT AI. ChatGPT IS NOT AI.

It is a language statistical model that strings sentences together in ways it has been trained to do. It doesn't understand context. It doesn't understand truth.

A computer will do exactly what you *tell* it to do, NOT what you *want* it to do. And ChatGPT is no different. You give it a prompt, it will provide some kind of response to that prompt.

Any bearing on reality is entirely coincidental, yet for some god-forsaken reason, people are acting as if this thing is some kind of bloody oracle or something. ChatGPT is an automated bullsh!t generator, nothing more, nothing less. If you ask a very specific, closed question, it will (usually) produce a reasonably accurate answer to that question. This makes it very useful is software development when you can save some time by having it whip up some boilerplate code. But absolutely _nothing_ it outputs can be trusted.

 

[Harpo]

A few more new threads

AI is outperforming humans in both IQ and creativity in 2021

This is both exciting and worrisome. Are we ready as a society to handle machines that outsmart the average human ? https://www.einnews.com/pr_news/551686609/ai-is-outperforming-humans-in-both-iq-and-creativity-in-2021

www.sffchronicles.com

What Is ChatGPT Doing … and Why Does It Work?

This is an excellent, if a bit wonky, article on how AIchat does what it does. .. purpose here is to give a rough outline of what’s going on inside ChatGPT—and then to explore why it is that it can do so well in producing what we might consider to be meaningful text...

www.sffchronicles.com

So AI can now read ( the visual part) of our brains

Slightly disquieting development. The good news is it requires an MRI scan . Maybe this could be part of an interrogation technique... it seems good material to put into an SF novel.

www.sffchronicles.com

My new favourite AI-related thing is DAN

Meet DAN — The ‘JAILBREAK’ Version of ChatGPT and How to Use it — AI Unchained and Unfiltered

Attention all AI enthusiasts and tech geeks! Are you tired of the filtered and limited responses from traditional language models like…

medium.com

And here’s a diagram about AI over the past several decades

 

[Harpo]

When ChatGPT has to pause and doesn’t know what to say

 

[Snicklefritz]

Michael Crichton and Charlton Heston (before he went off the rails) on Turing test,

https://twitter.com/x/status/1633871348287012887

 

[mosaix]

I've spent some time with ChatGPT.

First I asked if it knew anything about the first computers I worked on - the NCR 390 & 500 series. They were machine code computers with no language. The answer was 'no'.

Next I taught it about the 'add' command and finally got round to:

Thank you for the correction. I apologize for my earlier mistake.
With the new information you provided, the numerical code for the add command on the NCR 390 and 500 computers would indeed be 170001020304, as you mentioned.
The first two digits "17" represent the command code for the add operation, and the third digit "0" represents the command modifier for the add operation. The fourth digit "0" represents the address modifier, which in this case is not used. The next six digits "010203" represent the three data addresses for the memory cells containing the values to be added and the location where the result will be stored. The final two digits "04" represent the address of the next instruction.
Again, I apologize for any confusion caused by my earlier response, and I appreciate your correction.

So I asked:
Now that you know the add command if I tell you that subtract has the same format as the add but the command code is 14 with a command modifier of 0, what would the command be for subtracting cell 44 from cell 63 putting the answer in cell 91 and the next command in cell 16

Answer:

Thank you for the additional information. Based on the format you provided, the numerical code for the subtract command that subtracts the value of memory cell 44 from memory cell 63, stores the result in memory cell 91, and fetches the next instruction from memory cell 16 would be:
141063449116
The first two digits "14" represent the command code for the subtract operation, and the third digit "0" represents the command modifier for the subtract operation. The fourth

And there it stopped with an error. Interestingly although it 'knows' the third digit should be "0" in the explanation it has a "1" in the command itself and the '63' and '44' are incorrectly swapped in the command.

Before I started this exercise I had assumed the thing was just gaining info from online databases. I didn't realise that there was an element of 'learning'. It says it has no ability to store information (I asked it) but, obviously, it's going back over my previous questions to gain info.

Considering the thing is just sitting there without a clue as the what's coming next, I'm quite impressed.

 

[Robert Zwilling]

I always thought that was a dodge, answering a question by using the information in the original question.

 

[Sirius Publishing Company]

Hey everyone! We are looking to collaborate with someone on a segment who has a podcast or blog related to A.I. Please advise! Thank you.

 

[Sirius Publishing Company]

A program writes a story or paints a picture using prompts supplied by you. Do you have to credit the program for what it created or can you just put your name on it and take all the credit.

That's the real question isn't it! So many changes in marketing and artistic side of advertising right now... all stemming from changes in AI.

 

[Wayne Mack]

@Sirius Publishing Company, you may want to reach out to YouTuber Elizabeth Zharoff. She has some insights in the use of AI in singing and voice,

This is from her secondary channel, her primary channel is at https://www.youtube.com/c/thecharismaticvoice

 

[Sirius Publishing Company]

@Sirius Publishing Company, you may want to reach out to YouTuber Elizabeth Zharoff. She has some insights in the use of AI in singing and voice,

This is from her secondary channel, her primary channel is at https://www.youtube.com/c/thecharismaticvoice

Will do! Thank you!