What benefits you gained from AI

[MySiteGuy]

I was commenting on images only because various parts of the image issues have already been addressed by the US Trademark and Copyright office.

To be clear, I don't believe integrity is about whether claims they have a copyright over a song, article, etc.. Ultimately the courts decide whether their claims hold water. To me, Integrity is being honest when asked if AI helped, what part or percent comes from it, when it's brought up in a legal claim.

 

[Mr Lucky]

To me, Integrity is being honest when asked if AI helped, what part or percent comes from it, when it's brought up in a legal claim.

I agree, but my point was about defining AI.

But also in copyright it isn't always about percentages.If someone copied my work (a) but changes it a bit (to b) , my lawyer is going to argue that (b) could not exist with out (a) therefore the claim for (b) is 0%.

Years ago my manager was publisher of a famous UK band who had hits in the 70s. A remixer did a remix using samples and (quite rightly) asked what percentage they could have for using the samples. My manager responded with the very simple "0%". What you did could not exist without my client's creativity. That was the end of the story because the remixer's lawyers knew they had no case to go to court with. The only defences for that kind of thing are parody and fair use, none of which applied in that case.

In some respects copyright is as simple as that. In other respects it isn't.

 

[Wing]

Reminds me of how Paul McCartney from The Beetles was playing around with the idea of releasing a couple new tracks using AI generated John Lennons voice. Ethics eventually persuaded him to not go down that path, but this technology is looking to be very much copyright issue galore for years to come (if in the wrong hands).

Again, overall for the actual big picture of human interaction and life… I don’t really see much Pros for AI. Just a lot of negatives.

 

[RobinHood]

I was commenting on images only because various parts of the image issues have already been addressed by the US Trademark and Copyright office.

The problem is that as far as I'm aware, most of the big models won't disclose what the source of all their training data is. So how can the copyright owner even claim their content was used in the training and therefore in the output?

It's still the wild west. And in once sense it's a shame, because as much as I empathise with copyright holders and 'original' content creators, and I agree they should be compensated if their content was used as training data, it's obvious that the models were so much better before they started adding all these guard rails and clauses and protections. They're worthless without vast amounts of training data. So many of the models are nerfed in so many areas now compared to 6 months ago, the original magic and power is being drained from them. as more and more regulation and rules get added.

You can argue for days about copyright and attribution etc. but if the models don't have good data then results are crap. How do you make everyone whole and keep a good product without knowing how valuable each bit of training data is? I don't know the answer to that. It's just a shame to see some of the power of the models so heavily reigned in recently, sometimes to the point of being totally useless and telling you to do the task yourself. Makes me wonder about the long term viability of what is clearly some of the most revolutionary technology in decades.

 

[80sDude]

The problem with that is you have zero copyright protection (at least in the USA) for AI generated images.

According to the AI this is 100% true but if you modify the image even slightly you might be able to do so. Anyway for me I don't care about copyrighting any of its images. It's just a fun hobby. If I was financially benefiting that's a different story.

As an AI language model, I don't retain copyrights to the content I generate, so you can't copyright the images I create. However, if you modify or add substantial creative input to the images I describe, you may be able to claim copyright for the modified version. It's always a good idea to consult with a legal expert for specific advice regarding copyright law and intellectual property rights.

 

[leebo]

In the past I’ve created a 2 Siri Shortcuts that would run every day when I go to work. The first runs when my phone connects to CarPlay while one would run when I get to work. Both of these checks several sensors or smart-home items around our house to make sure everything at the house was ok. I leave for work at 2am and often I’m dead tired and forget to shut off some lights or forget to check to make sure some area of the house is OK.

The CarPlay one quickly checks anything major (water sensors, door sensors, and so forth) to make sure I’ve handled the major stuff before I leave the house. If there’s anything major it will give me a heads up right away to go deal with the issue.

The second one runs automatically when I get to work and really checks over the entire house. Items like lights, alarms, sensors, and so forth are all checked. Many of these items I can deal with remotely so there’s not a major need to deal with them in the car……but I’d still like to address the issues promptly.

In the past the messages Siri would read me more or less a script as it said the same thing over and over, which was fine, but repetitive. This weekend however I redid some of the shortcut to take all of the readings from all these items and combine it with weather and events from my calendar. All this data was then sent to AI where all the data was compiled into a summary of my house, daily weather, and daily events……all automatically.

Good afternoon! Let's start with your home: the alarm is currently switched off, and there’s an urgent issue that needs your attention—a leak under the sink has left water on the floor. It's important to address this soon to prevent any damage.

As for the weather, it's 47°F and cloudy right now. The forecast doesn't predict any significant rain today, with the temperature expected to drop to as low as 30°F by early morning. While the chance of precipitation slightly increases around midnight, it's minimal, and no substantial rain is expected during your primary hours outdoors (3am to 1pm).

You also have a couple of events today: "Little flowers" at 2:30 PM and "Sophia dance" at 5:00 PM. Make sure to plan your day accordingly to fit these activities in.

Stay dry, warm, and don't forget to handle the sink issue when you can. Have a great day ahead!

None of this is groundbreaking but being able to create a unique message daily that gives a snapshot is rather beneficial to myself. I’m really looking forward to when webhooks come to Xenforo (and hopefully include stats) so I can do the same with forum data. Picture getting a unique report daily that merges Xenforo data like new users or new posts with Google Analytic data like pageviews and search data…….all in a unique message created by AI.

 

[RobinHood]

Yesterday I subscribed to midjourney to test it out and just learn how it all works a bit better and tbh, It's absolutely worth the $10 starter subscription just for the entertainment value. I've laughed so hard with friends at some of the stuff that I've been able to create.

Something I'm experimenting with at the moment though is their blend feature, which I didn't realise it had, but is really useful, whereby I can upload my own content and manipulate it. This can produce results I'd never in a million years be able to recreate myself in photoshop, even though I'm pretty comfortable with it. And even if I wanted a human to produce it, I'm not sure I could. I know an insanely talented photoshop artist, and I don't think I'd even know where to begin in trying to describe or spec out the job to recreate some of the stuff I've been able to create. And even if I could describe or spec out the job, the cost for someone as skilled as them to produce what midjourney produces would be way out of reach for 95% of jobs.

Granted I'm talking about incredibly open ended experimentation and generation. If if the spec was highly specific, I'm sure a human artist or at least a considerable amount of additional photoshop post processing in addition to the ai image(s) would be needed. But it's such a fun tool to brainstorm and get the creative juices flowing.

 

[KensonPlays]

I only use it for small things, like creating outlines for blogs and whatnot. I don't have it write the actual content, I do that myself. Just the outlines, and title ideas. I'll usually tweak the title ideas afterwards, but only for basic stuff, the 'skeleton' of a post, but the actual 'meat' of it is written by me.

 

[Suzanne O]

I'll never use it.
You may not realise it but you're inviting the AI bots onto your forum. They'll eventually hack your forum and use your details.

 

[KensonPlays]

I'll never use it.
You may not realise it but you're inviting the AI bots onto your forum. They'll eventually hack your forum and use your details.

Eh? How so?

I personally use triple layer security for my admin area. HtAccess, 2fa, password. Should be pretty secure myself.

 

[Suzanne O]

Eh? How so?

I personally use triple layer security for my admin area. HtAccess, 2fa, password. Should be pretty secure myself.

AI is criminal activity waiting to happen.
You get AI bots on your forum. They are willing to break the law and hack your forum.
I've seen it happen. I'm not letting it happen to my forum.

Hype vs. Reality: AI in the Cybercriminal Underground - Security News

This report discusses the state of generative artificial intelligence (AI) in the cybercriminal underground: how cybercriminals are using ChatGPT, how they're adding ChatGPT features to their criminal products, and how they’re trying to remove censorship to ask ChatGPT anything.

www.trendmicro.com

 

[zappaDPJ]

I'll never use it.
You may not realise it but you're inviting the AI bots onto your forum. They'll eventually hack your forum and use your details.

You'll almost certainly want to use it to protect your forums from the very threats you are currently concerned about.

For example risk-based authentication employs AI to build user login profiles which offers a layer of protection that goes beyond multi-factor authentication. It already provides a pretty solid method of identifying and rejecting a login attempt to a compromised account.

It's inevitable that cyber-criminals will employ AI to gain the advantage so it's essential that software developers do the same in order to try and stay one step ahead.

 

[Suzanne O]

You'll almost certainly want to use it to protect your forums from the very threats you are currently concerned about.

For example risk-based authentication employs AI to build user login profiles which offers a layer of protection that goes beyond multi-factor authentication. It already provides a pretty solid method of identifying and rejecting a login attempt to a compromised account.

It's inevitable that cyber-criminals will employ AI to gain the advantage so it's essential that software developers do the same in order to try and stay one step ahead.

No thanks i don't use things that grab your details for identity theft. I've had enough of that criminal activity thanks.
YOu don't realise that they will also copy your voice and then go ringing everyone else up so they pretend it is you.
Perhaps think about that first.

 

[imno007]

Check out these video samples created with AI:

Sora: Creating video from text

openai.com

 

[Dad.]

Yeah this era is nuts. Its game over for text-based forums (that do not innovate), it already was. Even video is being scheduled. If you are not doing something truly interesting with your businesses, business in general could be going to the wayside over the next decade. Just a nutty time to witness.

 

[EllieCat]

Check out these video samples created with AI:

Sora: Creating video from text

openai.com

Imagine creating movies by just writing a single paragraph or sentence? In the future, we can also add dialogue and sound effects.

 

[Suzanne O]

Imagine creating movies by just writing a single paragraph or sentence? In the future, we can also add dialogue and sound effects.

Then imagine with all the info you give up that it comes back to bite you on your backside when they start using your details for identity theft.

 

[RobinHood]

Reddit just signed a deal to licence it's content to Google's AI tools for $60M per year.

https://www.reuters.com/technology/reddit-ai-content-licensing-deal-with-google-sources-say-2024-02-22/

Could this be signs of a potential lifeline for active independent forum owners that AI content licencing deals could on the cards if their content is good enough?

Or will the big players just scrape all the forum content as they have in the past?

 

[Suzanne O]

Reddit just signed a deal to licence it's content to Google's AI tools for $60M per year.

https://www.reuters.com/technology/reddit-ai-content-licensing-deal-with-google-sources-say-2024-02-22/

Could this be signs of a potential lifeline for active independent forum owners that AI content licencing deals could on the cards if their content is good enough?

Or will the big players just scrape all the forum content as they have in the past?

Reddit is the worst social media platform out there. It's overmoderated and utter garbage.

You sure you really want AI running your forum as well as pinching your details?

First-Ever AI Fraud Case Steals Money by Impersonating CEO

The first AI fraud case involves the head of a company who supposedly contacted the CEO and requested a transfer of funds. Read more about deepfake crime.

www.idtheftcenter.org

Do some real research on it.