More intelligent than you. Probably

[Justin Passing]

If you haven't had a go, try ChatGPT , https://chat.openai.com/chat

You have to register with email and phone, but it makes for a fun answerbot. It's good where you don't quite know what to look for, etc.
It's not very up to date (couple of years I think) and it doesn't know everything.
I asked it hard questions, and it can give different answers depending on god knows what. I asked it to derive an equation and it first went in circles, then hung, then it got it exactly right. If you ask it something with a long answer ( I tried derive Maxwell's equations) it does get there but you have to take it in stages.

The heading is tongue in cheek, but the thing does often "frame" an answer with things I wouldn't have considered.

 

[Koolpc]

If you haven't had a go, try ChatGPT , https://chat.openai.com/chat

You have to register with email and phone, but it makes for a fun answerbot. It's good where you don't quite know what to look for, etc.
It's not very up to date (couple of years I think) and it doesn't know everything.
I asked it hard questions, and it can give different answers depending on god knows what. I asked it to derive an equation and it first went in circles, then hung, then it got it exactly right. If you ask it something with a long answer ( I tried derive Maxwell's equations) it does get there but you have to take it in stages.

I would rather watch paint dry...

 

[Odds]

It's going to become a huge problem trying to sift exam papers in future.

 

[robinbanks]

At some point they'll have an AI do it

 

[Old Salt]

It's going to become a huge problem trying to sift exam papers in future.

Get them to sit in a classroom with invigilators and pen and paper. No electronic devices allowed.

There’s a novel idea.

 

[robinbanks]

I would rather watch paint dry...

I find it genuinely incredibly helpful for my work, but it doesn't involve drying paint!

 

[Justin Passing]

Exam papers --

At some point they'll have an AI do it

They already have! This adds complexity to that.

Get them to sit in a classroom with invigilators and pen and paper. No electronic devices allowed.

That's not so useful in many situations, people generally have time to look things up. Doctors in Emergency depts use systems like this to help at the edge of their knowledge. Granted you don't give the students time to look everything up...

I would rather watch paint dry...

Ah, a man who realises he doesn't need to know anything. Lots of jobs for chaps like you, in politics. PR dept.

 

[Dork Lard]

I wonder if they'll allow it to tell us lies?

On the basis that it is the information that is 'fed' to us that largely forms our opinions & influences our decisions, & that information can be manipulated & influenced by whoever controls the source, then once we've placed our 'trust' in the results from the questions we ask AI, then AI will largely control our opinions & influence our decisions.

Are you scared yet? I'm not.

 

[ETAF]

i have played around with it for a while now- good fun

not sure about lies ,
But ift does gets things wrong quite often... I have probably had 3 or things wrong in the few time i have played with it for factual stuff (not recent , as it only goes up to Oct 2021 )

I have also played with DALL-E - which is an Art version , also very good fun

I have used a few times for Excel (which i answer questions on a few forums) just to see what it suggested and it does well for intermediate formulas , and i read someone had used for code VBA and it produced a good version which just needed a little editing to get perfect

I saw an announcement yesterday that Microsoft is including in office 365 , which looks quite powerful , including teams for business.

Microsoft 365 Copilot Will Add AI to Word, Outlook, and More

Microsoft is rushing to add AI features to its services, such as the new Bing Chat AI and a new sidebar in Microsoft Edge. To no one’s surprise, AI functionality is now coming to Office apps.

www.howtogeek.com

Not sure when it will be released to personal 365 uses - but looking forward to playing with that when its released.

They have released another version, where images can be entered , the article suggested you can photograph ingredients and it would return a recipe.

I feel sorry for any teachers, as this will be the way to produce all sorts of work - summary of books, 500word articles on historical events and synopsis of a film , letter of complaint - all of which i have tried , and had some good results
maybe even replace some journalists

I asked it how i would tell the difference between an AI output and a Human, and it offered some info about humans we include more feelings and emotions.

I first saw it work on Question time a while back - where it was asked to write an opening for the show in the style of Shakespeare

 

[blup]

Google offers results based on the wording of the search, now it amends the search to help you understand what you meant.

Blup

 

[ETAF]

yep, google (well alphabet) lost a lot of money because of an error in their advert for the AI - Bard

Google shares lose more than $100 billion after AI chatbot Bard flubs answer in ad

The tech giant posted a short GIF video of Bard in action via Twitter, describing the chatbot as a "launchpad for curiosity" that would help simplify complex topics, but it delivered an inaccurate answer that was spotted just hours before the launch event for Bard in Paris.

economictimes.indiatimes.com

 

[Justin Passing]

I wonder if they'll allow it to tell us lies?

On the basis that it is the information that is 'fed' to us that largely forms our opinions & influences our decisions, & that information can be manipulated & influenced by whoever controls the source, then once we've placed our 'trust' in the results from the questions we ask AI, then AI will largely control our opinions & influence our decisions.

Are you scared yet? I'm not.

Not yet, but I agree the potential is there.

It has opinions.

Somewhat related, I ran across a video yesterday that I found thought-provoking. It is by Sabine Hossenfelder, a German physicist whose philosophy I like.

"I believe chatbots understand part of what they say. Let me explain."

 

[Odds]

Not yet, but I agree the potential is there.

It has opinions.

Somewhat related, I ran across a video yesterday that I found thought-provoking. It is by Sabine Hossenfelder, a German physicist whose philosophy I like.

"I believe chatbots understand part of what they say. Let me explain."

"Believing something is true doesn't necessarily make it so."
(Jean-Luc Picard. Captain; U.S.S. Enterprise.)

 

[Dork Lard]

I'm following a project where they're getting it to code a website from scratch, then populating it with content . . . . .

How good will it be when we can sit down & get it to design a building . . . . 'Cos the top ten list of people I hate in the whole wide world includes architects.

 

[ETAF]

'Cos the top ten list of people I hate in the whole wide world includes architects.

but doesnt use the historic info on the interent as a "learning" thing and so will combine all those architects and come up with some monstrous