General Discussion

highplainsdem

(58,457 posts)

42. You're welcome! I hope his experiment gets a lot of attention.

Sun May 28, 2023, 12:18 PM

May 2023

Sometimes errors even in public demos aren't caught.

Google got hammered by bad publicity and its stock lost value when it first rolled out Bard AI and its hallucinations were caught immediately.

Microsoft got VERY lucky with its demo of Bing AI, because that chatbot also made mistakes, hallucinated, during the demo, but those mistakes weren't caught till later, and Bing was hyped tremendously...and then it went off the rails and its ability to respond had to be sharply restricted.

I'm sure these AI are giving a lot more incorrect or simply crazy results than we ever hear about. Students using them for cheating aren't likely to call attention to that.

And businesses using fallible AI won't publicize that, either - won't want people to know about hallucinations/errors - for fear that it will damage their reputation.

Edit history

Please sign in to view edit histories.

Recommendations

0 members have recommended this reply (displayed in chronological order):

47 replies

= new reply since forum marked as read

Highlight:

College professor had students grade ChatGPT-generated essays. All 63 essays had hallucinated errors [View all] highplainsdem May 2023 OP

They're everywhere. marble falls May 2023 #1

" I'm much more worried about the possibility of us reverting to where AI is." WestMichRad May 2023 #2

In grad school (UCLA), I tutored math in Beverly Hills... Lucky Luciano May 2023 #7

As a fellow math tutor, I commend your technique. . . . nt Bernardo de La Paz May 2023 #9

Good Job ProfessorGAC May 2023 #10

I think the difficulty for most of us is in "context switching". When we're interacting with a tool erronis May 2023 #12

The best way to explain the productivity destruction of context switching! Lucky Luciano May 2023 #24

Very nice depiction of the tangled mess we weave. I think a lot of current software erronis May 2023 #32

Did the Wendy's drive-thru, I handed $20 on an $18.60 order and got $7.40 back. Returned it. TheBlackAdder May 2023 #30

The correct change is $1.40. You got $7.40 back. Could the cashier have calculated it on a progree May 2023 #31

The Wendys kid probably thought Danascot May 2023 #34

Hardly surprising. plimsoll May 2023 #3

Kind of reminds me of the old chess computers...... essaynnc May 2023 #11

I think you are right - Garry Kasparov was the Grand Master. And he is now knows a lot about AI. erronis May 2023 #13

prognosticating... oioioi May 2023 #17

The question is: Will be best at what? plimsoll May 2023 #18

ChatGPT is trained on the content of the entire internet. Yavin4 May 2023 #4

Not really. Only linked pages. There's a lot of content that isn't accessed without erronis May 2023 #14

The future of Chat bots won't be trained on available datasets like the internet. Yavin4 May 2023 #19

Agree. And am sure we don't know where this will be going. erronis May 2023 #21

I believe you are correct intrepidity May 2023 #23

Agree, AI has important uses as you describe, radius777 May 2023 #33

The question we need to ask is what GPT will be like thre years from now. speak easy May 2023 #5

They will go proprietary. Yavin4 May 2023 #20

We can regulate proprietary entities speak easy May 2023 #26

You could ask that question on most message boards now. n/t Yavin4 May 2023 #28

OK. Well how about the easiest and most effective way to build a bomb, speak easy May 2023 #39

You can find that information out as well right now. Yavin4 May 2023 #44

Ai can save you time speak easy May 2023 #45

It probably will get better, but it could also stall out on improving accuracy... Silent3 May 2023 #35

AI is useful as a data dump, a starting point bucolic_frolic May 2023 #6

But the other search engines don't know what they haven't been fed either. erronis May 2023 #16

That was a great exercise by that professor Warpy May 2023 #8

Sounds like me. In the 1970s we knew how to generate reasonable gibberish. erronis May 2023 #15

There's a radio station near me where women call in their "mom fails". milestogo May 2023 #22

This level of AI is basically brand new Takket May 2023 #25

Maximum likelihood estimation perhaps. Lucky Luciano May 2023 #29

Children think it is okay just to make things up. speak easy May 2023 #36

This is a machine, not a child. highplainsdem May 2023 #38

It is a model of intelligence that mimics a child. speak easy May 2023 #40

I'm not entirely opposed to AI. But LLM AI by design are likely to highplainsdem May 2023 #41

"LLM AI by design are likely to hallucinate" speak easy May 2023 #43

See this: highplainsdem May 2023 #46

Right. OK - I think we are on the same page. speak easy May 2023 #47

Google highplainsdem May 2023 #37

His students are impressively thoughtful. Thanks for this, HPD Hekate May 2023 #27

You're welcome! I hope his experiment gets a lot of attention. highplainsdem May 2023 #42