Artificial Intelligence: How to Avoid

Hello,

I am anti-artificial intelligence. It's becoming increasingly impossible to avoid, and I'm wondering what schemes you all have come up with for how to deal with this brave new world.

People I talk to think I'm a luddite for avoiding it, which I find strange because it is so new and so dangerous. I feel the world dumbing down as the robots get smarter. Make the robot do it! It can go FASTER!

Forget faster, I want accuracy and human knowledge.

I'm trying to dump google/gmail because they were the latest in my world to admit to me that their services are just AI bots cloaked in the familiar brand outfit. I need to kick my Facebook habit, but it's addicting on purpose. It was designed to keep me on there.

What are some services you use online that are still AI free?

Hi Carmen,

This is a huge rabbit hole to go down. I'll try to give some consise ideas that don't answer your whole question.

With any online service that's free, you have to ask why it's free. The answer is either benevolence, or that you/your data is the product.

It tends to be that if you pay for a thing, the features added usually pertain to what you want. And some places will add AI features, paid or unpaid.

So I would look for email and services that you pay for. There's a huge range of options. tuta.io (free + paid), Posteo, Mythic Beasts, and numerous others.

That said, you can roll your own. But that's an even deeper rabbit hole.

One of the problems with where we're at now is the alure of free. Many people would rather pay $0 to Google and have their data sold, than $1-3/month to another company that doesn't sell your data.

I would advise getting your own domain, however, so you can move your email with you if a paid provider starts implementing AI or other things you don't like.

-Theodorin

I had some more thoughts that I wanted to write down. This is more for the admins.

One issue with AI/machine learning is that corporations will crawl your content and then use it to generate answers. For now, they sometimes use sources, but it doesn't mean that your content isn't used as a secondary source.

One thing that you can do to help with this is block known crawlers from your site. This is actually pretty easy to do.

I see that Permies isn't currently doing this: https://permies.com/robots.txt

But it's pretty trivial to do: https://github.com/ai-robots-txt/ai.robots.txt/

It won't stop it all, but it's a good start in my opinion.

Thank you for these ideas.

I think what I'm gathering is that I do not know enough about the internet/websites to be able to avoid the AI bots, and that the only way out is to remove myself more fully than I have. I don't have the time to learn how to build a website so I can use the internet the way it was meant to be. This is not a complaint, but an acknowledgement of my own reality.

It's all a journey towards wholeness, over and over and over again until we meet God.

I'm hoping that the staff team sees your post and chooses to implement these resources that you've suggested. If not, that's okay. It's not my website I'll still use Permies anyway. The next time I'm online, I'll try to find a way forward with the email client issue.

I'm ambivalent about AI. It can have tremendous utility, but at the same time it's poisoning the internet commons, taking humanity even more abstraction layers away from physical ground truth.

Practical advice: you can avoid any sort of search function or auto-recommendation on sites like youtube. As you say, *the goal* of these sites is to keep you addicted, and *as a side effect* they occasionally show you high quality things. That's the right way to think about it. Instead of randomly surfing, find people (online and IRL) who you trust. When you hear of a video or channel that's high quality, get the specific thing you came to see, in and out. Don't engage with the site's inbuilt functionality at all.

I wish I had the discipline to follow my own advice. I've found some wonderful things on youtube that were random algorithmic recommendations. I've also wasted countless hours of time...

What a wonderful rabbit hole.

It helps to define what we mean by Artificial Intelligence (ai) or Machine Learning as it's called in the trade.  

AI has been a part of the internet and computing for over 20 years.  All search engines are AI driven as are spellcheckers, and anything else that learns from human behaviour.   I first became aware of programs that data mine websites without permission in the winter of 1996-97.  So that's been around for ages too.  

Or are we talking about the new advancements in AI like Help Chatbots (part of the internet for over 10 years now)?

Or the really new ones that made the media over the last couple of months like ChatGDP or the image generators.  Sadly, I can't see any huge leap in this technology as it's been around for so long.  Just some people put them together in a new and exciting way and a shiny new price tag.  

Image and text generators have a long way to go before they replace human skills.  And, apparently, there are now a whole flood of new job offerings for people who can generate prompts and sort through the results to get the desired ends.  Talking with people on the front lines, it appears many of the bigger companies are spending more on prompt generators than on real human artists - and are rehiring the artists they let go at a much higher rate.  

As someone who cannot communicate in writing without the aid of a spellchecker, I'm very glad to take advantage of AI all these years.  The newer versions don't help me as they anticipate my needs instead of following my lead.  But that's been a problem ever since some 'genius' invented autocorrect... so 15 years or so?

AI is like any tool - it depends on how the person uses it.  We can use fire to cook a meal or to burn down a city.  But I wouldn't want to live without fire.  

As for permies, my understanding is we use permaculture principles to approach bots.

The thing to remember is that without bots, there is no internet.  Many bots are good.  some bots are bad.  Most don't matter.  So like bacteria.  

The sledge hammer approach to blocking bots is like hitting the body with high dose broad spectrum antibiotics.  Kills good, bad, and indifferent bacteria... and too much of that kills the human.  We need bacteria to live.  

A website needs bots to live.

Block all bots, and we block search engines which means no more traffic to permies which means no more ability to infect brains with permaculture and share a better way with the world.

And if we are seen to block bots, then the good bots go away and the bad bots just go around and do their thing anyway.

So a more nuanced approach is needed and that is the level of detail I'm at liberty to go into here.



As for avoiding AI mining content - everything on a public site is mined.  Avoid spellchecker software and search engines as these areas have been at it the longest.  Read the terms of service before use.  Practice good internet hygiene.  Don't post anything publicly that you don't want to be known by the world.  If it's locked behind a login screen, it's less likely to be mined by external agents (we have a special PIE forum like this that google and the bots cannot see).

Personally, I would love to expand the secret PIE area to include new forums - but I'm waiting to see how people use what we have now to decide what direction to go with this project... does that make me a data miner if I'm a human doing the learning?  



We also have this option: https://permies.com/t/153455/activate-stealth-mode

The more people share about what AI is and isn't, I've realized that it's not AI that is the issue, but the replacement of human input in the internet commons that really gets to me.

Spell check, search engine suggestions, etc. are not my issue. When I see art or read a post, I want to be able to believe that I am connecting with someone's spirit. That is just not the case anymore.

AI bots are reading my emails and sharing that with corporations and then selling the content of them to advertisers. It happens so quickly, that I'm starting to believe that I don't even have to send the email before my words are sold to the highest bidder. Those words belong to me, the words I'm typing now are a collaboration between me and Paul because he owns the site.

I know the name of the man who owns this website, and can correspond with him. That's cool. If I fork over a few hundred bucks, I can go to his house and learn stuff. If AI helps him do that, that's pretty cool.

I can't even call the phone company without speaking to an AI chat bot and having them feed my every word to an advertiser. Who owns Verizon and how much do they know about me? Who are these people who are profiting off of my time and attention? Would the owner of Verizon teach me how to have a more meaningful life while using my cell phone? NO! They want me to waste my life on their product and spend more $$$

Why can't I speak to my doctor anymore? Why can't I call someone for help? Why do I have to speak to the robot first? The robots don't even know what they're doing!

I can understand that this is so much bigger than I realize, but I think what people miss is that it's so big that we are losing out on person-to-person relationships. Spell check and search suggestions aren't the problem. Those are tools, not content. AI generated content is dangerous and diminishes the value of human relationships. Children have stopped learning because they can't outpace the bots, so they have the bots do their homework for them. I did not have these resources when I was in school 10 years ago. It's going too fast.

Yeah, one big issue with these bots is that they are mandatory in so many cases. I don't mind to speak with a machine if that makes it easier for me to get an answer in time, but I want it to be optional.

Also, people love to learn just for the pleasure of learning. You could say that, since no human person is able to beat the machine playing chess, then interest for chess should have diminished. However, with so many chess bot teachers, children are playing more than ever and learning at a faster pace. My kid is playing regularly in chess.com, where he can find opponents of his same level, which is more fun than playing in the chess club where he is consistently beaten, or with his classmates where he always win.

I tend to avoid most things on the net.  I do have a handful of safe sites I use. While AI is a concern for me, so are many other things net related.

I would like the words I read, or those issued by an interlocutor, or video, or whatever, to be high quality (accurate, applicable, artistic). I don't care so much whether they're rolled up by some automation vs. produced directly by a human mind. Right now, these so-called AI tools are pretty bad, but in ten years, I expect they'll be better than talking to most people, and that's going to be awesome. This thing where everything is garbage right now is just growing pains and hopefully we'll get through it without destroying civilization.

I like what r rason wrote, the bots are neither good nor bad, its the intent behind the bots that can cause some problems.  And indeed they have been around for a long time.

That is why on things like the Y-tube thingy, I will several times each week enter some weird search words, often in another language, just to keep the little snooping bots guessing.  Unlike most folk these days I seldom go anywhere and when I do I wear a hat, glasses, face mask and undo my hair ( or hairs) so that it mostly covers my ears.  Some of those AI face recognition systems will check your ears for a match. (did you know peoples ears  are unique?).

As a non-conformist I feel a duty to enter as much false information into each system as often as possible (when legal).  And when I sell this house and move to my off grid forever home, I will not be online.  Just when on those few trips per year into the nearest town.

I really don't worry too much about "sky net" stuff.

Peace

I’m with you on creative content. I don’t want it mined by AI for wealthy corps.
One company has invented a cool tool to prevent this, a “skin” that visual artists can download and apply invisibly to artwork posted online. It’s called Glaze. It was developed at the University of Chicago. What it does is make the image appear as mud, a confusing puddle of texture and lines that don’t form any kind of image, to all AI programs and bots.

https://glaze.cs.uchicago.edu/