Urban75 Home About Offline BrixtonBuzz Contact

Artificial Intelligence Developments (ChatGPT etc)

Do you think AI is going to cause a technology revolution or cause economic destruction?
You’re in a desert walking along in the sand when all of the sudden you look down, and you see a tortoise, it’s crawling toward you. You reach down, you flip the tortoise over on its back. The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over, but it can’t, not without your help. But you’re not helping. Why is that?
 
You’re in a desert walking along in the sand when all of the sudden you look down, and you see a tortoise, it’s crawling toward you. You reach down, you flip the tortoise over on its back. The tortoise lays on its back, its belly baking in the hot sun, beating its legs trying to turn itself over, but it can’t, not without your help. But you’re not helping. Why is that?
Probably because I'd be a robot and wouldn't understand.
 
OpenAI Employees Say Firm's Chief Scientist Has Been Making Strange Spiritual Claims

A key character in the spectacle has been OpenAI chief scientist and board member Ilya Sutskever — who, according to The Atlantic, likes to burn effigies and lead ritualistic chants at the company — and appears to have been one of the main drivers behind Altman's ousting.
[...]
"Feel the AGI! Feel the AGI!" employees reportedly chanted, per The Atlantic, a refrain that was led by Sutskever himself.

The chief scientist even commissioned a wooden effigy to represent an "unaligned" AI that works against the interest of humanity, only to set it on fire.
 
Sports Illustrated Published Articles by Fake, AI-Generated Writers

There was nothing in Drew Ortiz's author biography at Sports Illustrated to suggest that he was anything other than human [... but ...] his profile photo on Sports Illustrated is for sale on a website that sells AI-generated headshots, where he's described as "neutral white young-adult male with short brown hair and blue eyes."
[...]
Making the whole thing even more dubious, these AI-generated personas are periodically scrubbed from existence in favor of new ones. [...] Each time an author was switched out, the posts they supposedly penned would be reattributed to the new persona, with no editor's note explaining the change in byline.
[...]
Though Sports Illustrated's AI-generated authors and their articles disappeared after we asked about them, similar operations appear to be alive and well elsewhere in [their owner] The Arena Group's portfolio.
 
I’ve been interviewing people at work for a cloud Architect role.
If you’re going to pump the design scenario into chatGPT and present the output at an interview, at least make sure you know what it all means.

“Under DR you’ve mentioned ‘pilot light’, can you tell me more about what that is?”
“Sorry, I left that in by accident.”

Same for “Circuit Breaker Pattern”.

🙈
 
There's a smart phone app used by many visually impaired people, (Be My Eyes) in which amongst * other things, you can either call someone from the Microsoft Accessibility answer desk or ask their AI assistant for help. It's actually really good and understands you maybe using a screenreader. i.e. is not answering with click on whatever.

have used a couple of times today already. Beats wading through Google results, cookies, ads and out of date info...

*. The other thing I most use it for is describing photos, which it does very well.
 
very true
im sure to be wrong
but to repeat my point, the genius of dance music is nailing uplifting basics hard, a much harder trick then people realise
being abstract is actually easier, easier still is being abstract and moody
i expect AI to nail abstract moody first - sweet spot electronic dance musis last :D
OK, this took less time than I thought. Check this out.

I didn't write the lyrics. Just used this prompt "A song about a guy called Ska who doesn't believe AI can make dance music".

It created the lyrics, song and video in under 2 minutes :bigeyes:

View attachment ska_ai_song.mp4
 
I made this for a mate. He loves Italo and he was banging on about some genre I'd told him about. Apparently he said I told him my boss loves a genre called Billy Boy Rock, and he knew I was bullshitting the whole time. I couldn't remember wtf he was talking about, but then eventually twigged he meant New Jack Swing (which my boss does love - no bullshit here!)

Anyway, I made this for my mate last night. It's probably the best one so far. Maybe because Italo is so cheesey anyway, it works?

View attachment billy_boy_rock.mp4
 
I wrote an article which ended up a bit tortuous and hard to read so I asked ChatGPT to make it more readable and it did improve it.

It isn't world domination by AI, but it helped me :) :(
 
Just as long as neither you nor anybody else on this thread is feeding it any personal data for your articles. Because that would be against European and UK data laws
 
Just as long as neither you nor anybody else on this thread is feeding it any personal data for your articles. Because that would be against European and UK data laws
I'm not sure that's been tested/proven

How is it differently from, say, using Google Docs and adding personal details to your document?

In both cases, you are transmitting personal data to a third party.
 
I'm not sure that's been tested/proven

How is it differently from, say, using Google Docs and adding personal details to your document?

In both cases, you are transmitting personal data to a third party.

I presumed he was being tongue in cheek because there is a very empty harbour somewhere.
 
Last edited:
I'm not sure that's been tested/proven

How is it differently from, say, using Google Docs and adding personal details to your document?

In both cases, you are transmitting personal data to a third party.
My understanding is this, which is based on conversations with our IT and data security head at work, whose job it is to care about these things. He knows what he’s talking about, but that doesn’t mean I haven’t picked up the wrong of the stick, of course.

The situation was tested with the Schrems II legal decision in 2020.


However, Google servers are in Europe and the UK for users connecting from those regions. The data doesn’t get taken to the US. And if you were working around that somehow (for example with a VPN) then you were breaking data regulations.

That’s similar to my own company’s situation. We have servers in the EU/UK and the US. The personal data we hold in the former is not allowed to travel to the latter.

In June 2023, the US and the EU signed a new data adequacy law, which allows data transfer where companies have proved they’ve jumped through certain hoops.


I’m not sure whether the likes of Google and Microsoft have jumped through those hoops, but I suspect so? Either way, though, I’m pretty sure that OpenAI have not. And if you think they have, the onus is on you to make sure before you start feeling it personal data.

There are additional difficulties with giving personal data to OpenAI too, of course. That data is being used to train the next iteration of ChatGPT, which is not a use the originators of the data signed up to.

[ETA] Oof. Looking around, it seems that OpenAI have been challenged with a whole bevvy of data protection breaches in the EU.


I wouldn’t personally be going anywhere near the thing with personal data until all those issues have been properly resolved.
 
Last edited:
You probably want to be wary about what you are storing in 'the cloud' lest it gets used for training AI. Dropbox admit that this may be the case if you enable any third-party AI services:
Anyone hoping to store their files/data securely in the cloud and keep them private, control access to said data, should do so using some independent intermediate encryption layer that they are happy to trust (eg perhaps CryFS, Veracrypt, Cryptomator).
 
Last edited:
Back
Top Bottom