AI - most astonishing reply

[Onholiday(somemightsay)]

no idea what the fuck this thread is about.
A1
Weren’t they a boy band?

@Mad Eyed Screamer - loves City, Morrissey & covered bridges (not necessarily in that order)........

Who taught you about boy bands? :-)

 

[JASR]

I just coincidentally googled a persons full name earlier this morning and the AI answer presented as the top result gave a potted history of the person with a timeline.

Great, except it combined the history of 2 very different people (with the same names) into the life of one.

My experience with AI chat bots used by companies on websites, inevitably leads me to type ‘speak to agent/human’ as the bot hasn’t got the answer/prompts for entirely irrelevant things related to my inputs/offers choices that don’t include ‘none of these’

I’m an IT person for 35 years, and so far all the times I’ve had to engage with AI, have been useless/wrong/time wasting.

 

[KS55]

It’s a road, heads oop north

No, it goes South to that there London.

 

[KS55]

Google is now experimenting with AI generated answers to queries. Often very wrong. I noticed it gave City’s debt at £millions and growing. So dangerous.

 

[denislawsbackheel]

When my iPad recently updated to offer this shite I turned it all off in settings, and Siri.

 

[Elbow beards]

Google is now experimenting with AI generated answers to queries. Often very wrong. I noticed it gave City’s debt at £millions and growing. So dangerous.

I googled whether my tv could handle dtx audio pass through. A.I. Said it could, lying ****, when on manufactures site who categorically said it couldn't.

 

[TonyM]

I just coincidentally googled a persons full name earlier this morning and the AI answer presented as the top result gave a potted history of the person with a timeline.

Great, except it combined the history of 2 very different people (with the same names) into the life of one.

My experience with AI chat bots used by companies on websites, inevitably leads me to type ‘speak to agent/human’ as the bot hasn’t got the answer/prompts for entirely irrelevant things related to my inputs/offers choices that don’t include ‘none of these’

I’m an IT person for 35 years, and so far all the times I’ve had to engage with AI, have been useless/wrong/time wasting.

It will take time for sure to reach the required level. I'd think 2 years for search results to be completely reliable. (But still dangerous due to the malicious risk of anything online)

 

[JASR]

It will take time for sure to reach the required level. I'd think 2 years for search results to be completely reliable. (But still dangerous due to the malicious risk of anything online)

It'll take longer than 2 years.
It's the same with any process - bad input data = bad output data.
To train the LLM theyve sucked, for want of a better way of describing it, random data from many sources. Those sources are generally the web - and all the misinformation that has on it. That information can slowly, manually, be corrected, but they still have to pull in new data, to keep the answers relevant/hopefully correct, so the amount of issues isn't going to decrease, it will increase, as the amount of data on the web continues to rapidly expand.

 

[shemnel]

The elephant in the room for LLM developers is that LLMs are now training themselves either on their own output, or output from other LLMs, which is proliferating the web. This creates hallucinations, results based on gibberish and a negative feedback loop.

 

[Man-chester copperpot]

Many of us have tried out A.I. for a while now, with ChatGPT being the most popular one.

Let's have a go at sharing what was the most astonishing experience or useful info you've received from your AI.
Or perhaps tip in general.