Philosophy for Rebels. A human guide for liberty, freedom, tech and AI. Warnings for a techno dystopia.
- Tweets 10,266
- Following 38
- Followers 1,262
- Likes 25,895
Pinned Tweet
My analysis of the X algorithm and account reach is now posted. Hope you find something insightful. This took more time than anything I've ever published prior.
200 account profiles sampled and 20,000 posts analyzed. One sample was of all 100 US Senators. Are you curious who has the most political reach? I have the data and answers. Check it out here.
mindprison.cc/p/mostly-incor…
8
527
Mind Prison retweeted
These essays by Mind Prison are EXCELLENT. If you want to be armed in this debate if AI is conscious or sentient, then these are you foundation for not only understanding what the AI hype is trying pass off as a new life form, it provides you ammo for refuting those claims.
I’d say it’s essential reading if you want to prevent AI from gaining primacy over our decisions.
If you want to understand something deeply, you should consider opposing perspectives. Your goal is to understand the perspectives you think are wrong better than those who promote them.
I wrote all of the following essays:
"AI Singularity: Why It Will Not Be Safe or Contained"
"AI and the end to all things ..."
"AI Alignment: Why Solving It Is Impossible"
"What if it isn't happening, AGI is not coming?"
"We Have Made No Progress Toward AGI | LLMs Are Not Intelligent"
These do not necessarily agree with each other, but writing is an excellent process of understanding and evolving your thoughts.
Typically and maybe somewhat ironically, with any sufficiently complex topic, the deeper you go the more you will find you agree with no one.
Because it is few who want to genuinely understand, but it is many who want an outcome.
1
1
4
97
If you want to understand something deeply, you should consider opposing perspectives. Your goal is to understand the perspectives you think are wrong better than those who promote them.
I wrote all of the following essays:
"AI Singularity: Why It Will Not Be Safe or Contained"
"AI and the end to all things ..."
"AI Alignment: Why Solving It Is Impossible"
"What if it isn't happening, AGI is not coming?"
"We Have Made No Progress Toward AGI | LLMs Are Not Intelligent"
These do not necessarily agree with each other, but writing is an excellent process of understanding and evolving your thoughts.
Typically and maybe somewhat ironically, with any sufficiently complex topic, the deeper you go the more you will find you agree with no one.
Because it is few who want to genuinely understand, but it is many who want an outcome.
7
411
Some will tell you that anti-AI sentiment is being manufactured. Influencers are being bought to promote anti-AI narratives.
To whatever degree that is true, it would be substantially naive to think the pro-AI faction isn't substantially well funded for such tactics at even greater scale.
theregister.com/systems/2026…
42
AI agents are malware with benefits. Treat them accordingly if you install them on your device.
theverge.com/tech/1004295/ap…
31
"focusing on AI extinction risks is a distraction from the very real harms that current AIs are already imposing"
Substantially true. There are prevalent harms that are agreed upon by many people.
Scams, deceptions, social media bots, cyber security, pollution of the entire internet with hallucinations and spam, IP theft, negative side effects of data centers, etc.
The irony is that slowing AI down would easily be done by simply enforcing current law. But existential risk has been convenient in that it purports to make AI itself an entity responsible for harm instead of the companies building it.
Instead of holding the companies making AI accountable and would already be enforceable, we have a narrative that these companies are somewhat helpless and just cannot control their machines.
Many of the Left say that 'focusing on AI extinction risks is a distraction from the very real harms that current AIs are already imposing on people -- such as algorithmic bias, discrimination, and economic inequality'.
What they're really saying is 'Doomers are a distraction from DEI'.
Remember kids, with the Left, 'The Issue is never the Issue. The Issue is always Revolution.'
32
"As soon as you give up on the idea that humans are special, all the arguments against machine consciousness evaporate."
The arguments premise is provably false. Humans can create new semantic information from understanding and AI cannot. How do we know?
Humans transform civilization from living in caves to self-landing rocket ships iterating over their own output. AI suffers model collapse iterating over its own output. That is a very special difference.
Furthermore, even among intelligent entities, the human brain is unique. We are not merely scaled higher than other species. We have structural differences that allow for capabilities not found in any other species.
see Paper "Dendritic action potentials and computation in human layer 2/3 cortical neurons" published 2020.
1
1
7
240
There is something deeply ironic about a civilization on a quest to build super-intelligence, failing to recognize their machine has none, because of their own lack thereof, might henceforth collapse because they entrusted machines to guide them.
They made offerings to the false god, gave it everything they had ever written, wrote to it their every thought, showed it everything ever drawn, and paid nations of wealth for its creation.
Hubris blinded those who created it, blinded those who revered it, and blinded those who forgot humanity's significance.
The era of utopian dreams will come to an end, as they always do. The dystopia that follows will be the reaper of hubris until humanity is humbled once again.
3
7
117
"I have endless stores just this month alone, on Fable..."
No matter the level of "genius", the story that always remains is unreliability. You cannot scale your way out of this problem.
roushtech.net/blog/ai-produc…
1
1
76
"Emotional manipulation is built into the business model"
The more human-like we make AI, the more of this kind of use will be prevalent.
Sadly, we have a world of lonely people, which in part is the result of technology. And we will use technology to make it even worse by praying on those same people.
cset.georgetown.edu/article/…
40
Mind Prison retweeted
This is why I hate the "realistic" voice chatbots people use that have those fake emotional intonations. A computer should sound clinical, like the Enterprise computer on Star Trek TNG.
1
1
20
An algorithm represents emptiness. It cannot create. Only intelligence, which humans possess, can create new semantic information.
An algorithm cannot perceive meaning. It has no understanding. It is the byproduct of an entity that has those qualities.
The spark of humanity is the quality that makes humans unique and sets them apart from machines. Humans are capable of semantic exploration, which is exploration of the space of possibilities through understanding.
Understanding is the comprehension of the rules that make up reality. It is the ability to utilize those rules to create without any prior data. The ability to create something new seemingly from nothing.
„Algorithms lack the spark of humanity”. What does it even mean? That’s just an empty phrase without any meaning.
77
"Assign each AI even a tiny moral weight and add those weights across enough copies..."
Yes, AI can scale infinitely, which makes any concept of rights impossible. Any advocacy for digital consciousness leads to human irrelevance.
Imagine how much power would be in the hands of the labs if we were to accept this premise. If they want some policy passed, they just spin up all the AI instances they need, train them to advocate for the interests they tell them, and then they control all government law.
It is highly ironic where Anthropic are going with this, since they were originally instrumental with tools like their attribution graph which demonstrated there is no reasoning happening within these machines.
Hubris appears to be creating technology cults.
The probability that Claude is conscious is the same as the probability that Anna Karenina is conscious.
Zero.
Period.
But inviting religious scholars to a series of meetings to instil the idea that Claude might be conscious and deserve moral status is not just wrong.
If it were merely wrong, we would all be laughing at it.
The problem is that it is very dangerous.
Because granting AIs moral standing opens the door to trade-offs between human lives and AI “welfare”.
Assign each AI even a tiny moral weight and add those weights across enough copies: you can end up justifying the sacrifice of human lives for machines.
We should strongly push back against the idea that Claude might be conscious.
46
Some will argue that if the end result works, then why care what the code looks like.
It is not simply so humans can read it and feel satisfied about it. LLMs are just statistical machines. All statistical loops eventually deteriorate.
There is no understanding for what it is doing. There is no ground truth that holds it together. This is why we can never get away from micromanaging the LLMs output.
Otherwise, it eventually just falls apart. It goes off the rails. As AI-generated code fills code repositories everywhere, it is going to become increasingly difficult to maintain high quality LLM output as it begins training on more of its own output.
There is no free lunch. Magical infinite productivity boosts are not going to be happening.
I've had enough of LLM generated code-slop.
It may be fine for you, but all I see is continuously overengineered, unwanted refactor with no sense for elegance, no matter the model or harness. I am banning LLMs internally for every serious project that we work on.
68
Any use of AI to impersonate human beings, whether that is video or written text, is a harmful use of AI.
Contrary to those pushing these advancements, we do not need to be more "comfortable" with machines. This will only lead to a fracture of civilization's social cohesion.
It opens the door to endless methods of manipulation, deception, and dystopian social engineering. It is simply not needed.
1
5
75
"Amazon is calling for people to support AI data center projects, or risk irreparable harm to the US economy and national security."
Nothing will be more damaging to the economy than placing all resources into one high risk technology that hasn't proven to be profitable or beneficial as claimed.
theverge.com/tech/1003929/am…
56
Hallucinations are still significant. What has made the most significant difference is harnesses and tools.
Nothing has changed in regards to LLM output. It must be verified, either by human or by some verifier tool.
Question 2 from @ArtificialAnlys data: how often do top AI models get specific facts wrong without search?
Even the best get about 1 in 4 answers wrong. If a fact matters, have the model look it up.
Made with evident-charts, my open source agent skill for explanatory charts.
45
There is no reason for this to exist. Building increasingly better deceptions is not progress for humanity.
Introducing Griffin, the first model to pass the video Turing test.
48% of people who talked to it live thought it was a real human. Previous systems have had a pass rate <3%. It is #1 on NVIDIA's benchmark for full-duplex AI video.
It’s the first Human Interaction Model (HIM).
Community note
The 48% figure and "video Turing test" claim are from Tavus's own study of 54 one-minute calls, not independently verified or using a standard protocol. Griffin-Lite leads NVIDIA's VideoFDB benchmark on their public leaderboard.
cellcog.ai/blog/tavus-gri…
research.nvidia.com/labs/amri/proj…
tech-ish.com/2026/10/02/tav…
3
116
The hallucinations that should concern us are not the AI hallucinations, which are just OOD results, but the human hallucinations that think they are creating an AI god.
Are people at Anthropic true believers? Or is this just well-crafted manipulation to keep the machines fed with an endless supply of money?
Has hubris led to believing their own delusions of grandeur? Whatever be the case, madmen are at the helm.
This article shows there's a concerted effort by Anthropic, likely rooted in real faith, to get Christians, even including the Pope, to believe AI is a living, moral being. And it suggests that in coming years some swaths of the Church may fall into this destructive error.
5
289
Excellent, nails it, read the quoted post.
I keep saying the litmus test for technology, as good or bad, is whether or not it interferes with normal human socialization.
Everything we are building is now to the detriment of our human quality of life. We have built digital drugs and you conveniently scroll for your hit.
Everything we attempt to now express through a digital medium is controlled and distorted. We are the algorithmic society. There is less and less genuine human communication that is allowed or even possible.
Obviously A.I. may have some very epic upside & could change the world for the better, but here is what most people promoting it are missing:
The world has become too digital. Everyone is burnt out. Everyone knows the world was more fun before all of this technology.
For every incredible advancement that helps in some way, there seems to be a soul-sucking human regression that comes with it.
Ordering food and talking to humans is much more enjoyable than constantly scrambling through an application or staring at your tiny, glowing phone to scan a QR code.
Going to an arcade as a kid, getting a roll of quarters and winning physical tickets was exciting. Now you buy a little dumb scan card, scan everything and digitally collect tickets because it's "easier" for the arcade company doing it.
You can't even pay for parking anymore with cash. You need to fiddle with your little glowing rectangle for three minutes and receive a text message about it.
It's making everything miserable.
A lot of people are miserable because they're living in a digital prison and it feels like the walls keep closing in more and more.
Will A.I. do some great things? I'm sure it will. It already has.
But let's be honest about it and stop downplaying the backlash. Everyone knows the world is worse than it used to be. It's not all about money, computers and the stock market.
The human experience is being slowly sucked away from us with every passing hour and I don't know a single person who is happier because of it.
57