Sponsorship

Saturday, April 15, 2023

One Of Our Most Brillliant Minds

 . . . of the 20th Century had already laid out a roadmap in 1942 that should have been adopted and passed into Law right there and then. We all knew it was coming

That brilliant mind belonged to Isaac Asimov and that roadmap was the Three Laws of Robotics

And why not? It would at least have given a framework. A place to start, a place to build a new set of Laws from. 

Asimov's Three Laws of Robotics

First Law

A robot may not injure a human being or, through inaction, allow a human being to come to harm.

Second Law

A robot must obey the orders given it by human beings except where such orders would conflict with the First Law.

Third Law

A robot must protect its own existence as long as such protection does not conflict with the First or Second Law.

Zero'th Law

A robot may not harm humanity, or, by inaction, allow humanity to come to harm.

(This last law is an addendum by I Asimov as it became necessary. Proving that all Laws need to be flexible and extensible.)

I'd extend that to include "any person, corporation, and/or AI that creates an AI or robot" and that all be held equally accountable for any breach of these laws. And also that the words "humans" or "humanity" be replaced by the words "beings," "entities," "species," or "ecosystems" as appropriate. 

How Would It Apply

If ChatGPT3.5 (or whatever) had been used to eliminate a person's job and that person decided to seek redress for that harm: The plaintiff would have to prove harm. They would have to prove the harm was a result of the robot(AI)'s activity. (Writing an article / drawing an image for your boss would constitute harm I imagine.)

The defendant would then be free to seek out whether it was their creation that caused the harm without any other cause, or whether some other person caused the robot(AI) to commit that harm.

If the AI approached the boss and claimed it could write better copy than the plaintiff that would be the AI's fault. If it could be proven that the third party caused the harm to occur by soliciting the product from the AI then the case becomes about the third party as the new defendant. 

So an acquaintance one of whose major clients just terminated their casual copywriting employment arrangements because they were now using GPT would be free to sue that client for the termination of the arrangement but would probably lose a suit based ony on that.

BUT if they sued OpenAI and OpenAI could prove that the malfeasance was down to that client, a malfeasance would have been proved, and THAT might make a difference in how companies and corporations use AI to cut staff. And look - in effect, the AI costs the company or corporation nothing, so even if they go with the copy from the AI, they should still retain the creatives as part of their duty of care.

And it's good to think that even as customers of those third parties, we can have some effect on their decisions. We can move our own accounts to a new provider that respects their employees. Look at this:

The above is an example of depersonalising you. And me. And all their other customers. We deserve to be seen as people, not these ridiculous caricatures. You've probably noticed these kinds of ads everywhere. I'm betting that most are now being made with AI image programs and the creatives that used to be behind the art have been let go. 

I can't get behind that. Yes, in the short term it may stop prices rising but they will, again and again. And in the meanwhile, it's put someone out of a job. So why should you and I support this? The only way I'd support it is if I knew the employees who were technically out of a job were still getting the income they deserved. 

And getting back to the new style of "clip art" being used in ads. Original clip art was a really crappy way to put a commercial graphic artist out of a job, and let others do the publication/ads work. I have a particular axe to grind here, yes. My father was a commercial artist/graphic artist and he (luckily) re-skilled anyway out of a desire to do bigger and better things, and I'm glad to say he never really had to deal with Microsoft's crappy blight on Powerpoint forever. (He was around, but he was only ever a home user of computers.)

The important takeaway from all this was that clip art depersonalised the graphic artist, and then became the burden of the poor office bod that had to make presentations and Powerpoints after the graphic artist had been sent on their way. 

But ads were still done by graphic artists. Right? Right? Well. Some. With some assistance of software effects. But these new graphics do look a bit AI generated to me. I use NightCafe and DALL-E2 to generate me some themed images that I then layer over and under and around to make stuff to fill whitespace on the blogs. But compare that muppet above ("Allegra Style" I think they call it?) to the unedited AI images I had AI create for me:


Miles better, and that wasn't even a complex prompt. (Also, if you know, you know - the top image was a DALL-E2 I think, the second from NightCafe. I think...) But I don't really remember. Also, I generally don't put the straight piece up, I prefer to edit the heck out of it first and make it more my own than someone else's secondhand style. 

PS: I'm not sure where I stand on artists' work being imitated. I deliberately try to avoid style directives in the prompt and hope my fairly simple instructions will do the trick. The above were generated from "anthropomorphic cat in a lab coat / holding a test tube / holding a laser" respectively. I couldn't hope to pay for a graphic artist to make those, and spend between one and six (!!!) hours on them. The above, I just downloaded and posted. 

And if you want to stop reading my blogs because I didn't employ a graphic artist well I'd understand that - but I also write my blog posts by myself (across six themed blogs I semi-regularly write to) without resorting to GPT - and I notice that very few readers bother to make a donation to me nor even a regular monthly stipend donation. So maybe you're part of this ongoing cheapness problem.  

I used to put a footer on my posts with links to a newsletter and a News Stand and several different ways people could make a one-time or repeating donation and so far I've been surprised at how hard it is to make any kind of dent in the outgoings. Which is - I think - one of the reasons I'm so anti what's being done to creatives right now in the name of economising. 

And it's also a reason I do think that every human is owed a Universal Basic Income from the profits that are currently being stashed away by every corporation, Bezos, Musk, and every corporation sitting on billions of dollars of dead money that'll never see circulation in the economy again (by way of paying their damn taxes) so - let's start agitating for that, especially now that corporations are exploiting machinery and AI to rob people of a livelihood. How about it? Are you in? 

Find my contact details (If you check the banner at the top of this post it links to donations - but also a place to read up and get involved - and the first step is to DO SOMETHING) and get in touch. Dare ya!