57 Comments
User's avatar
ToxSec's avatar

🌍 ☠️

Expand full comment
chris kalaboukis's avatar

Actually, the Internet’s been dead for a while. Only now with AI are they livening it up a bit

Expand full comment
ToxSec's avatar

You actually bring up an interesting point. If we say the internet’s been dead for a few years, maybe in a few years from now when we can’t tell the difference between people and LLMs it might feel more alive than ever. Imagine one that drops really good memes tailored to your humor hah.

Expand full comment
chris kalaboukis's avatar

i think we are already there. The internet is both more dead and more alive than ever. Alive with bots, that is.

Expand full comment
ToxSec's avatar

For sure. I saw the reports recently. More than 50% of the traffic. 80% on some sites.

Expand full comment
chris kalaboukis's avatar

Is Reddit 80%

Expand full comment
ToxSec's avatar

I haven’t seen official Reddit stats, 80% was X based on a July 2025 paper.

Expand full comment
chris kalaboukis's avatar

X is for sure 80%+ bots

Expand full comment
ToxSec's avatar

Company estimates: Twitter has stated that less than 5% of its monetizable daily active users are bots. 😂😬😬😬

Expand full comment
chris kalaboukis's avatar

You believe that? When was the last time you were on X?

Expand full comment
Gavin's avatar

Already we have ai powered browsers like Atlas and Comet summarising and guiding us to content.

While I would like to see more governance and regulation of AI content - AI is moving faster than laws can keep up. Like you mentioned our best defense is mindset.

Expand full comment
ToxSec's avatar

That’s a fantastic point. Ai browsers will be summarizing and guiding. Will we lose nuance? What about a subject OpenAI doesn’t like, will Atlas down play it? Leave it out entirely?

Expand full comment
Gavin's avatar

Will be very interesting what comes out of this.

Control of people's searches and sources.

Would governments try to regulate this?

We are living in some interesting times.

Expand full comment
ToxSec's avatar

100%. You give up some control. OpenAI shows their demo of adding sunblock to your shopping cart when you ask it. Which does it choose? Who owns the company that contracts with OpenAI? I wonder if their brand gets clicked by the agent over the one who isn’t associated.

Expand full comment
Gavin's avatar

I can see ad revenue and payment for improving a business search ranking becoming a thing for OpenAI and other vendors.

Expand full comment
ToxSec's avatar

Yeah, I think it’s just a matter of time a this point.

Expand full comment
Dallas Payne's avatar

This is fascinating to follow while simultaneously really horrifying! I am so curious about the conversations we'll have about this a few years from now and what would have happened between now and then. We'll also have a whole new host of concerns... I'm already doubtful of most things I find in a google search as I know how much I can create with AI at just my regular user level.

Expand full comment
ToxSec's avatar

I agree! 2025 will be a turning point I think. The way we think of pre-Facebook internet. I will have even less trust on what I find online lol. Now I don’t trust I’m talking to a person for the first time!

Expand full comment
Gregory Brenton's avatar

Think back to the good ole days. When a troll could simply ignore was all that cluttered a good exchange. Or the twenty-somethings that wore funny hats in public and took pictures of their food, and we thought "how strange". Suppose there is no going back.

Expand full comment
ToxSec's avatar

I know we've had 'bots' for a while, but it's crazy how fast I got nostalgic for the pre ai internet..

Expand full comment
Chris #TheAntiVirusGuy Moody's avatar

Will this kill off social media as we knew it? Will social media turn into a "fantasy media" or "artificial media" or "bot media"?

It's already appearing to be happening to LinkedIn. Will LinkedIn implode as we get more and more "bots" shouting at each other, drowning out the humans?

Expand full comment
ToxSec's avatar

I’ve had such a difficult time curating a LinkedIn feed that sounds human and now just people shouting and sharing their new cert. bots everywhere it all sounds the same.

Let’s hope we can find innovative ways to solve, I really like Substack and would hate to see it follow the same path.

Expand full comment
Chris #TheAntiVirusGuy Moody's avatar

I moved over to here three months ago because my LinkedIn feed was constantly full of utter rubbish. Hopefully Substack will stay fairly “bot” free.

Expand full comment
ToxSec's avatar

Fully agree. I force myself to do it.. it’s like eating broccoli haha. Idk how much longer I will though…

Expand full comment
Chris #TheAntiVirusGuy Moody's avatar

What I need to find out, is a way to get some of my LinkedIn followers to subscribe to my Substack.

If I could just get 1% or 2% of my LinkedIn followers over to here it would be brilliant.

But they don’t seem to want to.

Expand full comment
ToxSec's avatar

Yeah, my LinkedIn is small, (1k) because it’s never been interesting to me. But I still post teaser. I’ve gotten a full 5 follows that way haha!

The are dug in it seems.

Expand full comment
Chris #TheAntiVirusGuy Moody's avatar

Yes, they all seem fiercely loyal to the platform @ToxSec

Expand full comment
Saxxon Creative's avatar

What I am enjoying is the hallucination a.i and how the dream sphere of peoples reality is becoming a hallucination proven by the ability to not tell the difference between sora 2 videos and the real videos. Server Tower of Babel moment.

Expand full comment
ToxSec's avatar

It’s trippy. I’ve also seen research comparing hallucinations to dreams when trying to understand how LLMs think.

I think we are surely entering a period soon where we have an authenticity crisis. We won’t be able to tell what’s real on electronic devices without serious new techniques. But then whoever controls the Authenticator, controls the truth!

Expand full comment
Saxxon Creative's avatar

Oh wow never thought of that.. That is brilliant observation. brings into focus the idea of who is the original… like red dwarf show.

The Robot only lies when no one else is watching… I think programmers of LLMs based most of there computer behaviours on Red Dwarf TBH.

https://www.youtube.com/watch?v=8525OKIhwqk

Expand full comment
ToxSec's avatar

Super interesting will take a look. Thanks!!

Expand full comment
The Threshold's avatar

Cannot agree more our best defence is user education and critical thinking. Not only because it is AI but because humans make mistakes too. I am all for progress and finding the right balance. Regulation can help, but make it too strict and we won’t be able to progress to deal with this issues. I hate seeing the limitations I already find within AI for a few bad players.

Expand full comment
ToxSec's avatar

I think you have a great mindset here. Balance is the right word. Too much or harsh and we are just swimming upstream. But not enough regulation and who knows where it takes us. Good call out!

Expand full comment
Alexandru Giboi's avatar

Interesting take. Just be very careful: running from the AI filter bubble to the human private chambers is not quite evolution. It’s jumping from the stirring pan into the fire.

Expand full comment
ToxSec's avatar

It’s definitely going to be a challenge. How do we get a diverse human environment these days? Tough. I think there might be just our evolutionary tribalism in play. Ai or not! Thanks for reading Alexandru :)

Expand full comment
Alexandru Giboi's avatar

Of course we have a tendency to hide ourselves inside of a group, but AI & social media really push this too far.

Expand full comment
AI Governance Lead ⚡'s avatar

Great read!

Expand full comment
ToxSec's avatar

Really appreciate it!!

Expand full comment
Sam Illingworth's avatar

Great article, thank you. Do you think that the introduction of a 'human watermark' could help? And as a side point AI-generated comments should just be banned on sight. Content - no problem at all. But those comments, oh man! 😢

Expand full comment
ToxSec's avatar

I’m glad I’m not the only one who hates them. I sat next to someone who would copy and paste content from an LLM to their comment.

I did read that Substack might actually count engagement time while typing, so maybe less effective that they think it’s.

Great question! I’m trying to get a post up about watermarking. (more than just the sora icon you can remove) hopefully I’ll have an informed response soon. I know there are technical challenges, and some good ideas but invasive with your privacy.

Expand full comment
Sam Illingworth's avatar

Cool. Will really look forward to that. I know some of the camera manufacturers are exploring this topic too. You might also enjoy this piece from Vogue (of all places!): https://www.vogue.com/video/watch/multimedia-forensics-when-digital-photography-meets-artificial-intelligence-photovogue-festival-2023-what-makes-us-human-image-in-the-age-of-ai

Expand full comment
Esme Kealoha Dudoit's avatar

Social Media is a failed experiment. It was created by Peter Thiel to control elections and it runs as planned. Palantir never misses. Return to Reddit and IRC, Matrix. Bluesky and more so Mastodon have the right idea. Build servers and moderate them.

Expand full comment
ToxSec's avatar

I do have a very new Bluesky. I’ve tried Mastodon because I love the idea but maybe haven’t found the right community yet.

Expand full comment
Esme Kealoha Dudoit's avatar

Bluesky and Mastodon takes time to build connections, as it should. The notion of accelerated online friendships is a lie of the enshittification process that Doctorow is outlining.

Expand full comment
ToxSec's avatar

Maybe I’ll have to spend a bit more time building that community aspect up. Worth a shot!

Expand full comment
NOSMH's avatar

Great points. The problem is that average humans are far too lazy and uninterested in working for accuracy. They would rather assume if they can read it that it is true.

“Someone else surely has fact checked this before it was allowed to be shown to me.”

Expand full comment
ToxSec's avatar

Absolutely. In fact the idea of fact checking is under fire. I watched a video on YouTube about a team trying to get to the bottom of several well accepted scientific facts, and they spent 6 months in a circular list of references. Mapped it out, everyone just referencing papers that referenced papers. It’s only getting worse from here when AI makes up half the references lol.

Expand full comment
NOSMH's avatar

With AI we will find a LOT of junk science that we thought was more or less decided. We are already seeing it medicine and physics.

Thankfully we will still find that the Earth is indeed round.

Expand full comment
ToxSec's avatar

On a personal level, I’ve had so many old wives tales dispelled lol. I learning a lot of junk science is great.

Expand full comment
Terod Naej's avatar

This is a powerful piece that triggers relevant questions and answers. For me AI is a ressource, a tool we have to use responsibly… and I think we should mostly focus on that instead of going at war against it. I mean think about, we’ve been using all kinds of tools and ressources to communicate, to express ourselves, to connect; from the simple note books to high-tech computers and apps… and even the ghost writing concept… Yes, real really matters but are we asking ourselves what is really real, or what really matters? Is it the people who creates and produce? Is it the ressources? Is it the outcome? And yeah, what about the consequences? The fact is, we are all mostly right even through our different points of view? We just need to learn to leave with it including the AI concept? AI is real and is part of this chaos we the people are constantly creating. And as chaos comes with opportunity, why not be smart about it instead of fearing it?

We got this. Hope we do.

Expand full comment
ToxSec's avatar

You bring up great points! And for sure, I’m in AI security, so I end up writing “against” it, even though I’m a huge fan of it and use it daily professionally. I look at a lot of my pieces and you’d think I hate it lol. To your point, I think of it like nuclear tech. I’m all for nuclear energy, and the general public is starting to shift back to that. Kyle Hill has some great points on the subject. But we can also misuse nuclear too easily. AI is just already in the hands of everyone hah. I do hope we learn to use it responsibly.

Expand full comment
Digital-Mark's avatar

That's not a theory anymore. 🤣

Expand full comment
ToxSec's avatar

Rip internet.

Expand full comment
Meenakshi NavamaniAvadaiappan's avatar

Tracing the sources to the same for the good 😊

Expand full comment
ToxSec's avatar

+1

Expand full comment