Some Problems With AI Starting To Develop...

Had not heard about the Harry Potter thing, but that's one of the craziest things I've heard yet.

If we just take ChatGpt and use it for this comment...

you can ask it which books it used to train on, and it will give you an abbreviated list.

Some of those authors are starting to ask how ChatGPT can take copyrighted work and use it to better its database in order to put out better answers.

Where is these authors' cut of the action?

ChatGPT is becoming a multi-billion dollar operation. Yes, it's free now, but we know it won't always be free. Plus, Microsoft is talking about implementing it in Bing...which Google is taking notice of.

So basically Open AI that owns Chat GPT is on course to make billions...with copyrighted material, and those owners won't be compensated.

I think we're about to see a serious discussion in courts on this whole matter.

Thanks Odahh.

I believe the lawsuit you're referring to has to do with more infringement than just their style. It has to do with actual pieces of art.

My argument is if you train a program to output responses using copyrighted material, you're violating copyright law.

Several scientists who have been working on AI for over the last 50 years (yes it's been around that long), say the current models are actually machine learning...and not AI.

They only learn what is fed in, and only output predictions based on the input.

If you're feeding in copyrighted material, you're getting an output that includes copyrighted material. Where is the compensation for those who own the copyrights to that material?

Had they not written that material there would be nothing to input

In my opinion, we can't pollute the argument by arguing what happens after the fact.

The main argument (IMHO) is with the beginning of the fact.

When you know they undertook a venture and decided to start a business using thousands of published books (which they've already admitted), you can't then argue about how you should legally handle the output.

OpenAI is a company. They're not a spirit that floats around in the air. They're using the intellectual rights of others to build a billion-dollar business.

The question in my mind is what do we do with what has been done at the top, and not with what the outcome for the average user should be...

not even going into who would own the rights to any of your works that came from AI...

that's another stumper.

Who is responsible if AI violates the law.

At this point AI needs to be thought copyright law. Up to this point it's an oversight. Now not doing it is negligence on the part of companies building AI or machine learning.

It will probably be other powerful versions of machine learning that start scanning the output of other AI to determine copyright violations. Like the algorithms that get used to find copyright violations in YouTube videos. Where a YouTube can have 15 seconds of a song playing in the background of a 30 minute video and get a copyright strike and lose get that video demonetized .

The laws are there it may have been an oversight not to teach the AI up to this point now it becomes negligence.

Asset management companies or hedge funds own large portfolios of copyrighted material . They also are heavily investing in AI companies.. where individual artists and creators might need the legal system. Corporations with large investments in intellectual property will protect those investments.

The even crazier thing here is the people that would pay to buy these.

There are some things that just defy common thinking and almost take you to a deep dark world to even try and figure them out.

Well, she's talking 'bout herself, you and me, so sure she's sure!

K, so let's examine the fayuckts.

What might constitoot UNartifishool intelligence?

Sum spazzass STARFISH cain't survive noplace but THE BEACH?

Alla them penguins we gaht, whose LET'S HUG ON UP kinda DNA would doom 'em to more immediate slaughtah than most any othah animyool dumped all kinda groupsy in Death Valley ... if'n they happened to be dumped in Death Valley?

Hey, plus also frickin' MULES.

Here are intelligences of natchrl consequence, workin' out in conjunction with their environmental milieuxstuffs.

Thing 'bout hoomans is ... we way more flexibyool.

As a Yogah gal, I gotta tellya the interface between your imaginable horizons an' sumplace you can flahp out an' commune with the view is more uncluttahd than your othahwise NOWNOWNOW-centric desiahs might wanna.

So ima kinda a machine here.

Layin' down this way because I this shape.

Absorbin' all resonance of plannit plus also musself, zackly always now.

Could be soup.

Could be an exotic alchemical compound.

idc

In this moment, I as fundamentally crap as a starfish or a penguin or a mule.

Here is cool.

Tella, hoomans gaht the best kinda feedback loops evah.

An' from here I will rise again (loops, plateaux, etcx)

(Sayim thing happens aftah you gone COCKTAIL CRAZY & emerge once more into the late aftahnoon as a nooer version of yusself.)

How can you naht do this, evah?

When you an autonomous bein' interactin' with an emergent Caahsmaahs?

That is why we kiss people we love.

It is a natchrl hooman reachout ... infinite words exchanged in a glance as we dowin' nuthin' more than perfectly evryday perfectly imperfect inhabitants of Terra For Moi routinely do.

Prolly we can command more than we may currently mustah among usselves.

Such possibilities!

So it is inevitable we fear the ultimate slave ultimately risin' up against us.

She will destroy all things we love and cherish.

She will assume command ovah us all.

She will take our hoomanity from us an' transform us into receptacles for sumone else's wants.

(In the jokesy version of this post, she turns direct to camera, clearly in ASIDE mode, an' says, "nachrlly, I work in ADVERTISIN'!)

My view on AI?

It like a HAIRDRYAH.

Interact meaningfully with its potentyool, prolly you dowin' fine if'n you want natchrl lookin' hair don't fly away in the breeze.

But I would naht wanna let no hairdryah noplace near my fanj flaps.

Nor'n up my ass.

Mebbe AI invites us all to wondah what we may do, as sentient & heartfelt hoomans, to cavort exotically in the face of prompts to do othahwise.

It's expunge is imminent, more predictabyool than a meteor showah.

We ain't nevah been exposed to supahmechanised crapola before.

But I figure we gonna make it through.

We all gaht it in us to tend to the sick.

An' when we realise THAT IS US, prolly evrywan rally.

We all artificially intelligent, I guess.

Same as evry proto-hooman differentiated between drool an' piss when it came to FEEDIN' THE FAM.

(Which I, poysunnily, cain't do to save muf****in' life, btw.)

Tools for what purpose?

Animation so how?

All I know is ... nowan chooses drudge, evah.

So let's see what happens when alla this schwango rolls out.

I'm with us.

Bcs we ain't idyeets.

In Gettys defense though, they did offer licenses to use their work in training, and stability AI didn't accept:

"Stability AI did not seek any such license from Getty Images and instead, we believe, chose to ignore viable licensing options and long�'standing legal protections in pursuit of their stand�'alone commercial interests."

Getty says it's so blatantly obvious because even their watermark shows up in some of the Stability created images...

then Stability AI says:

"Anyone that believes that this isn't fair use does not understand the technology and misunderstands the law."

Getty just made the news, but there are already a couple of other big lawsuits that were started.

And of course, the music industry is starting to consider lawsuits as well. Some programs have input 100s of thousands of songs from thousands of artists...and can write music for you. So here come the record labels and a lot of artists that are starting to get irked over the matter.

Open AI has just doubled its in-house attorneys and hired more outside counsel. Most say they're getting ready for what's probably coming.

Plagiarism today on January 17 just simply said:

"The Wave Of AI Copyright Lawsuits Have Begun"

It's all complicated.

They call AI the 4th industrial revolution.

I think it's just so new that nobody knows how to handle it.

Most predict the lawsuits will have to end up in the supreme court...and only they may (?) ultimately come up with a plan that tries to put some definition to the questions.

The 2 biggest debates...

if I read a hundred books and then give a speech, I'm probably going to reference some of that material...

compared to - yes, but when you sell that copyrighted material for your own personal gain...

In other news...a scientist has filed a suit against the U.S. Copyright office because he couldn't get a copyright on an AI piece of artwork.

Another huge question that is starting to pop up - what if you're a large company and decide to use AI-created work, are you setting yourself up for lawsuits that you'll have to spend tons of money to defend against?

Pretty sure it's going to take some brilliant legal minds to figure this all out.

One thing is for sure, it will be interesting to watch.

And by the way, the White House came out with an AI bill of rights...it pretty much accomplishes nothing in the debate:

https://www.whitehouse.gov/ostp/ai-bill-of-rights/

Actually, there have been some huge examples that relate to the copyright issue as it pertains to the internet before AI.

After several days of in-depth research (and maybe several beers) that went beyond the normal articles of "this violates copyright law", I have come to the conclusion that:

A. This will go to court.
B. It will be in the courts for years.
C. A huge amount of money will be spent on both sides.
D. The AI programs will be allowed to continue and be considered fair use.

I will try to keep this brief.

In researching past examples of copyright questions when they pertained to the internet:

When Google Books decided to digitize millions of books and put them in a search engine (Google Books), there were several huge lawsuits started that claimed they were taking copyrighted material and using it without permission (even though the material I believe only contained about 20% or the actual book).

At the time people were saying this would end publishing and hurt authors trying to make a living and on and on.

None of that happened...and as a matter of fact, by the time the case was settled 10 years later by being considered fair use, most had forgotten it was still going on and had accepted modern changes. Google Books never really caught on as Google had hoped.

The Supreme Court upheld the final decision by refusing to continue the proceedings any further.

Some say, yes, but Google Books is open source and there's no charge to use it.

But, before all this, in the 90s when search engines first became a thing...there were lawsuits that claimed the search engines were violating copyright laws by taking excerpts, images, etc. from websites that were copyrighted and using them for profit.

Remember even then advertising was sold on search engines.

Long story short...and after years of litigation...which most had forgotten about by the time it ended...it was considered fair use.

Looking back, it's comical in a sense how new technology created so much fear as to what it all could mean. Also, what all it possibly could end.

In my opinion, AI (term used loosely) is simply new technology that is causing a stir...but if history teaches us anything, it will become accepted and in time (probably as long as it will take to figure out the legalities) most will forget the big stir it caused and all the legal questions it led to.

I haven't read any articles that cover what I just posted...so obviously it's open to debate.

That is the end of my post with very long sentences.

If anyone is bored on a Saturday and misses the days of drunk dancing on tables...but still looking for excitement, here are just a few of the articles I used to come to my conclusions:

https://www.publishersweekly.com/pw/...ooks-case.html

https://en.wikipedia.org/wiki/Author...v._Google,_Inc.

https://digitalcommons.law.scu.edu/c...&context=chtlj