Artificial Intelligence (AI) is progressing at warp speed - A bit nerdy but I think you'll be amazed
9 watchers
May 2024
2:21pm, 24 May 2024
50,081 posts
|
HappyG(rrr)
bbc.co.uk Google new version AI makes a few rubbish mistakes!  G
|
May 2024
2:36pm, 24 May 2024
21,535 posts
|
Chrisull
A human being can be more than just their influences or what they've been told (and saying people just parrot what they're told is an example of asymmetrical , an LLM can never be. I've just opted out on Facebook/Meta from them using my data to train their AI, it's buried in your notifications under, we've changed the way we use data, you have to email them an objection and they will decide IF they want to honour it. They've honoured mine anyway. I think in the future there's a very big class action coming against one of the big players over AI and plagiarism |
|
May 2024
2:37pm, 24 May 2024
21,536 posts
|
Chrisull
Attempted to tag a user in my post - and it just posted my post unfinished instead.
|
|
May 2024
2:39pm, 24 May 2024
21,537 posts
|
Chrisull
"asymmetrical insight". was the missing word. Was also attempting to say: cackleberry - the kid picking 12 and the computer not being able to pick a number outside 1-10 is a GREAT example. Exactly this. |
May 2024
2:57pm, 24 May 2024
17,127 posts
|
jda
If you asked a computer for a number between 1 and 10 and it gave 12 you would laugh at it for being broken, not praise it for being cheeky and intelligent. The random number argument doesn’t seem very relevant or even correct to me. Computers can certainly generate numbers that are random enough for any practical use. Humans are much much worse at this task, they really can’t do it at all (other than by copying the computer’s calculations by hand). |
May 2024
3:01pm, 24 May 2024
19,289 posts
|
Dave W
I agree. I don’t think that computers would have a “favourite number” that might skew their randomness. I think that the random number example was probably introduced to try to make it more understandable for non computer people like me. So thank you for that. But it just makes it more confusing. |
|
May 2024
3:38pm, 24 May 2024
50,083 posts
|
HappyG(rrr)
That's more of an example of a "spontaneous joke" that a computer can't do. But if it is not answering the question, being cheeky, that's not behaviour you want from a computer. Or not until we get to them being 100% correct and then we want to "humanise" them somewhat and introduce some humour and some truth filtering. Nowhere near that (see TARS and CASE the robots in Interstellar with their humour and truth settings. tvtropes.org ) G
|
May 2024
3:42pm, 24 May 2024
436 posts
|
DaveG
The example of a number was more about the thinking process, rather than the result. Computers will give numbers spread much more evenly than people, so is a good approach. But they need to retrieve it from someone, not magically produce the number. The example of a child saying 12 is precisely why this is important. If the question is 'How do we put telephone lines into inaccessible places?' a computer can explain the physics and technology needed to plan that. It can't suggest 'Is there some way of creating a mobile signal and handheld phones' if they don't already exist. That's the right answer, and you need the ability to say '12' as a number between 1 and 10 to have the intelligence needed to make those types of breakthroughs. Until a computer can do that, it will always be massively constrained and able to do things quicker, but not better, than humans. |
|
May 2024
6:34pm, 24 May 2024
19,296 posts
|
Dave W
So does true AI actually exist yet or is all that we currently have just supercomputers that are capable of pulling information from everywhere and passing that off as “thinking”?
|
|
May 2024
7:45pm, 24 May 2024
50,087 posts
|
HappyG(rrr)
That DW. But... is "simulating" intelligence - sifting lots of data, giving outputs that are based on that, usually factually right and sounding natural, and sometime creating unique stuff (the images, sounds etc that you'll have seen and heard) significantly different? As the robot in I, Robot answered to the question "Can a robot write a symphony or paint a work of art?", "Can you?". Cos I can't! It's getting close. And it needs serious design and governance. Imho. G
|
Related Threads
-
Rise of the robots :-)
Aug 2024
-
The ChatGPT action figure image generator
Apr 2025
-
chatgpt for coaching
Apr 2025
-
Night sky aka Star gazing aka Astronomy :-)
Jun 2025
-
Science
Apr 2025
-
What's your view on someone who believes Earth is being or has been visited by UFOs/UAP's
Dec 2024
-
Interesting study
Jul 2024
-
Units of measurements
Feb 2022
-
Mars Landing
Feb 2021
-
No limit to the benefits of exercise in reducing the risk of cardiovascular disease
Jan 2021
Report This Content
You can report any content you believe to be unsafe. Please let me know why you believe this content is unsafe by choosing a category below.
Thank you for your report. The content will be assessed as soon as possible.