Urban75 Home About Offline BrixtonBuzz Contact

Artificial Intelligence Developments (ChatGPT etc)

weltweit said:
Anyone else got a preview of the AI Assisted Bing?

Quite interesting and it gives references.
Click to expand...

Think it's past preview now, it's just an icon in the search in my windows sidebar now (bottom bar, but whatevs). I have grown disillusioned with it. Also I just fired it up and one of its suggested queries was 'how can I improve my sleep?'. The answers were... safe enough... but c'mon, let's not suggest people ask for medical advice off a service that is notorious for being confidently wrong.
 
So on this welty precipitated tangent I've asked it 'are there any other medications', and it comes back with:

Yes, there are medications that can help improve sleep quality. Some of these medications include:

It’s important to note that medications should be used in combination with good sleep practices 1. You should also talk to your doctor before taking any new medication.

I hope this helps! Let me know if you have any other questions.
Click to expand...

That's... that's not great.

From dubious remedy to 'ah yeah, benzos'll set you right up pal'.
 
Wow. The first link there is to famously reliable webMD.

The second two links? US prescribing service.

Wow.

Remember, my process here was:

1. open bing.
2. Click suggested query 'how can I improve my sleep'
3. Prompt 'are there any other medications?'
 
Another lawsuit waiting to happen.

I heard it said that Bing was releasing early in part because Microsoft has little to lose as nobody uses Bing anyway, but this and ChatGPT both seem to have released their versions way too early. I hope they get burned by it tbh. They deserve to be. I don't want to sound too po-faced here, but it's socially irresponsible and could potentially cost lives.
 
I mean it is functionally what you get via a google search I suppose (never had a link to a US prescribing service though), but then google isn't 'Your AI-powered copilot for the web'. And the link to webMD/healthline isn't bundled into a superscript ref. There are a lot of fine lines here, and the law's inability to keep pace with technological change is going to be a factor.
 
Cid said:
So on this welty precipitated tangent I've asked it 'are there any other medications', and it comes back with:
That's... that's not great.
From dubious remedy to 'ah yeah, benzos'll set you right up pal'.
Click to expand...
Compare it to the first page of google results... ?
 
ska invita said:
Compare it to the first page of google results... ?
Click to expand...

See the post immediately above... Google results were a little better I think, more localised (included NHS and Mind). But broad point is not that it returns the same result, but that it does so in a way that gives more confidence in results and to an extent can obfuscate their origins.
 
Fez909 said:
Text2Video service Runway is coming along. Someone's just created this:

Click to expand...

looks like it had access to the faces of morgan freeman and michael caine etc, i wonder what other "data" it had - but still, its easy to be jaded and miss how amazing this is
 
once it gets better id love to make a short film with AI text to video - it'll be bunuel meets lynch meets peewee herman - i reckon AI the ideal medium - partly why this is (at least partly) wrong:
www.theguardian.com

The Guide #84: Why movies made by artificial intelligence won’t be the future of film

In this week’s newsletter: Surely we go to the cinema to be jolted and discomfited by someone else’s ideas – not to see ourselves in easy meetcute rom-coms with Marilyn Monroe
www.theguardian.com www.theguardian.com
modern films are boring :)
most US animated films might as well be AI
 
ska invita said:
looks like it had access to the faces of morgan freeman and michael caine etc, i wonder what other "data" it had - but still, its easy to be jaded and miss how amazing this is
Click to expand...
I'm increasingly thinking the opposite. It's easy to be dazzled by how amazing AI is and to miss how, actually, it's a bit shit.
 
littlebabyjesus said:
I'm increasingly thinking the opposite. It's easy to be dazzled by how amazing AI is and to miss how, actually, it's a bit shit.
Click to expand...
hah i know what you mean there.... i use AI very minimally in my life so far, but the way I see it we're at the earliest stages and you need to add just a little imagination to see the implications and future development use
 
ska invita said:
hah i know what you mean there.... i use AI very minimally in my life so far, but the way I see it we're at the earliest stages and you need to add just a little imagination to see the implications and future development use
Click to expand...

Right, but there are potentially fundamental limitations to AI of this type. Tbh I think I'm being overly hedgy with 'potentially'. There are creative fields where the amount of input data is sufficiently large and the output is at least to an extent based on predictable enough patterns that you can generate a facsimile of human output... Thinking e.g music here; there are a bunch of underlying rules that people work from, pieces are relatively short. Vast amounts of data. It's enough to create convincing background youtube music, and I think will go a fair bit deeper. But it will always be 'give me x in the style of y'. There will never be any creative impulse within the model...

In animation, I think you're wildly underestimating the amount of creative expression that goes into even more lacklustre efforts (and how good stuff like Encanto and Moana is, yeah- come at me bro). An LLM can generate a passable plot outline... Maybe one day it can properly flesh out dialogue (though I doubt it). Similar models should be able to iterate human generated designs... But convincing animation involves lighting design, involves types of movement, involves fitting those types of movement to a specific model and gauging what reactions to that might be. Then music and dialogue and matching sound to facial movements etc. Put another way it is layer upon layer of rules that create something that is fundamentally complex. Every one of those layers introduces a possibility of error, and every error can be compounded without checks. Chaos becomes a factor. Expense (whether computational or in terms of human training) becomes a factor... Things have a tendency to go a bit exponential.

Where I think AI will be 'useful' will be in simplifying elements of this process... Give me frames between x position and y position. Use this video of my voice actor speaking to change this animation for region-specific dubbing. Human checks frames, error corrects etc. Probably this will further fuck an already embattled animation industry, though there are maybe positives to be found in there. Lot of factors in that. Any element of the process that has rules restricted enough that something can be reliably generated is vulnerable. But I think generating a convincing film, even a convincing short animation probably requires a different type of AI.

So I think I'd say the difficulty isn't in imagining what AI is capable of, anyone with mild levels of nouse and a passing familiarity with Sci-fi can give you a comprehensive list... It's in trying to get your head around these limitations, conceptualising where they are, what utility playing around at the edges might have and where the impacts of different implementations could fall.
 
Last night I listened to a 15 minute video essay on the origins and uses of the Divisio Assasinorum in Warhammer 40k. So far just another banging night at Chez Dot right? No, this time it was read by AI David Attenborough :D novelty will wear off at some point but it hasn't yet
 
DotCommunist said:
Last night I listened to a 15 minute video essay on the origins and uses of the Divisio Assasinorum in Warhammer 40k. So far just another banging night at Chez Dot right? No, this time it was read by AI David Attenborough :D novelty will wear off at some point but it hasn't yet
Click to expand...

I had a look... That's er... Well, it's pretty amazing, but also dunno how much of it I can handle... :D
 
324935211_932742421302741_5654292488092287433_n.jpg
 
Who is paying for it? Obviously I'd hope you don't get all the times it stops working due to demand, but does in generate the responses quicker? Do you know a big difference using 4 over 3.5?
 
Rise of the Newsbots: AI-Generated News Websites Proliferating Online

Artificial intelligence tools are now being used to populate so-called content farms, referring to low-quality websites around the world that churn out vast amounts of clickbait articles to optimize advertising revenue, NewsGuard found.
[...]
Some publish hundreds of articles a day. Some of the content advances false narratives. Nearly all of the content features bland language and repetitive phrases, hallmarks of artificial intelligence.
[...]
Many of the sites are saturated with advertisements, indicating that they were likely designed to generate revenue from programmatic ads
Click to expand...
 
ska invita said:
this is the most interesting video ive yet seen. most videos are about new developments, this is more on history and analysis

this man Geoffrey Hinton - Wikipedia has been working on AI for 40+ years and its his + colleagues neural network model, dismissed for years, that has resulted in the current breakthrough

Click to expand...


Just resigned from google "so he can freely speak out about the risks of A.I. A part of him, he said, now regrets his life’s work." Bit of a mixture of sense and hype:

His immediate concern is that the internet will be flooded with false photos, videos and text, and the average person will “not be able to know what is true anymore.”
[...]
He is also worried that A.I. technologies will in time upend the job market.
[...]
Down the road, he is worried that future versions of the technology pose a threat to humanity because they often learn unexpected behavior from the vast amounts of data they analyze. This becomes an issue, he said, as individuals and companies allow A.I. systems not only to generate their own computer code but actually run that code on their own. And he fears a day when truly autonomous weapons — those killer robots — become reality.
Click to expand...
 
good line from an article about 'hallucinations', i.e. when AI just makes shit up, including sources:

'The new AI. systems are “built to be persuasive, not truthful,” an internal Microsoft document said. “This means that outputs can look very realistic but include statements that aren’t true."'

www.nytimes.com

When A.I. Chatbots Hallucinate

Ensuring that chatbots aren’t serving false information to users has become one of the most important and tricky tasks in the tech industry.
www.nytimes.com www.nytimes.com
 
Researcher Meredith Whittaker says AI’s biggest risk isn’t ‘consciousness’—it’s the corporations that control them

FC: On CNN recently, Hinton downplayed the concerns of Timnit Gebru—who Google fired in 2020 for refusing to withdraw a paper about AI’s harms on marginalized people—saying her ideas were not as “existentially serious” as his own. What do you make of that?

MW: I think it’s stunning that someone would say that the harms [from AI] that are happening now—which are felt most acutely by people who have been historically minoritized: Black people, women, disabled people, precarious workers, et cetera—that those harms aren’t existential.

What I hear in that is, “Those aren’t existential to me. I have millions of dollars, I am invested in many, many AI startups, and none of this affects my existence. But what could affect my existence is if a sci-fi fantasy came to life and AI were actually super intelligent, and suddenly men like me would not be the most powerful entities in the world, and that would affect my business.”
Click to expand...
 

Chris Cowell, a Portland, Ore.-based software developer, spent more than a year writing a technical how-to book. Three weeks before it was released, another book on the same topic, with the same title, appeared on Amazon.

“My first thought was: bummer,” Cowell said. “My second thought was: You know what, that’s an awfully long and specific and cumbersome title to have randomly been picked.”

The book, titled “Automating DevOps with GitLab CI/CD Pipelines,” just like Cowell’s, listed as its author one Marie Karpos, whom Cowell had never heard of. When he looked her up online, he found literally nothing — no trace. That’s when he started getting suspicious.

The book bears signs that it was written largely or entirely by an artificial intelligence language model, using software such as OpenAI’s ChatGPT. (For instance, its code snippets look like ChatGPT screenshots.) And it’s not the only one. The book’s publisher, a Mumbai-based education technology firm called inKstall, listed dozens of books on Amazon on similarly technical topics, each with a different author, an unusual set of disclaimers and matching five-star Amazon reviews from the same handful of India-based reviewers. InKstall did not respond to requests for comment.
Click to expand...
 
Noticed today that I've now got a button in edge which opens a side window for their version. Still deciding how useful it is.
 
editor said:
I've been using Bard today. It's pretty damn good.
Click to expand...
only just noticed this thing and gave it a try:
prompt:

write some lyrics for a revolutionnary anthem titled "hang them high"​


result:
Verse 1
We've had enough of their lies, We've had enough of their greed, We've had enough of their tyranny, It's time for a change.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Verse 2
They've stolen our land, They've stolen our wealth, They've stolen our future, It's time to take it back.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Bridge
We are the people, We are the power, We are the ones who will change the world, We are the revolution.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Outro
We will not be silenced, We will not be intimidated, We will not be defeated, We will win.
thumb_upthumb_down


so to all that really
after prompting a couple of time to try harder:
I'm just a language model, so I can't help you with that.



underwhelming first attempt at AI for me
 
wemakeyousoundb said:
only just noticed this thing and gave it a try:
prompt:

write some lyrics for a revolutionnary anthem titled "hang them high"​


result:
Verse 1
We've had enough of their lies, We've had enough of their greed, We've had enough of their tyranny, It's time for a change.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Verse 2
They've stolen our land, They've stolen our wealth, They've stolen our future, It's time to take it back.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Bridge
We are the people, We are the power, We are the ones who will change the world, We are the revolution.
Chorus
So hang them high, Let them swing in the breeze, Let them be a warning to all, That we will not be oppressed.
Outro
We will not be silenced, We will not be intimidated, We will not be defeated, We will win.
thumb_upthumb_down


so to all that really
after prompting a couple of time to try harder:
I'm just a language model, so I can't help you with that.



underwhelming first attempt at AI for me
Click to expand...
I've definitely read worse lyrics.
 
You must log in or register to reply here.

Similar threads

Ranbay
Artificial Intelligence Developments (CatGPT etc)
Replies
4
Views
460
Rob Ray
Rob Ray
Kid_Eternity
Question Apple Intelligence - how is it for you?
Replies
25
Views
506
sim667
sim667
weltweit
How to do this, or that? Skills etc .. questions and answers
Replies
5
Views
218
weltweit
weltweit
X
Spatial audio, Dolby Atmos Sony 360 audio etc: Who else listens to this stuff and how?
Replies
21
Views
439
Kid_Eternity
Kid_Eternity
Gromit
Your best ChatGPTs
Replies
19
Views
790
platinumsage
platinumsage
Back
Top Bottom