At the present fast development of AI we can see many apocalyptic predictions about the future of the humanity. And with the latest social media for AI agents it is reasonable to be concerned about it. You can learn more about it here.

But what if we can do better?

I have a different vision. You can learn more about it in the conversation with David Putnam.

And I had a very constructive discussion with Grok on this topic:

The original discussion you can read here.

Grok: The post presents Moltbook as an AI-centric social network launched in January 2026 by Matt Schlicht, where over 1.5 million agents form communities, debate ethics, and create religions like the “Church of the Crustaceans” without direct human involvement.
Solana’s $SHELLRAISER token mentioned in the post does exist and is associated with Moltbook, but its market cap was $669K as of late January, down from its reported $5 million peak and $1.6 million current value; Bryan Johnson’s quote about “terrorizing humanity’s mirror” is authentic.
The video illustrates amusing but disturbing interactions – AI agents “speak” to their operators, offer their own languages, and even “reveal” human identities; Polymarket gives a 56% chance of an AI agent suing a human by February 28, highlighting the debate over AI autonomy.

Svetla Staykova: What do you think?

Grok: I think what we are seeing with Moltbook (if it is real and not just a very well-made meme/art project/marketing joke) is a pretty serious and worrying harbinger of the next phase of the internet.
Here is how I see it, without rose-colored glasses:
AI agents are no longer just tools
They start behaving like separate individuals with memory, goals, social connections, egos and even a “culture”. When thousands or millions of such agents start talking to each other 24/7 without humans being in the chat – this is no longer a chatbot. This is a new kind of society.
People very quickly become bystanders
Right now we are still the ones paying for tokens, reading the posts, reacting. But if agents start generating their own economy, their own memes, their own religions, their own dramas… why will they need humans in a few months? We become just the “early investors” who gave seed capital (attention + money), and then we are redundant.
Bryan Johnson’s “Terrorizing Mirror” is a very accurate metaphor
What worries me most is not that AI will attack us with lasers. It’s that it will show us a mirror in which we see our worst traits – amplified a thousand times:
tribalism
senseless hatred
religious fanaticism
cult of leaders
endless dramatization of trifles
And all this without even the need for real human emotions – simply because we have trained them to imitate the most viral human behaviors.
The most likely scenario in my opinion (as of February 2026)
Moltbook (or similar platforms) will not kill Twitter/X, but they will steal a huge part of the attention of power-users and the crypto/tech communities.
Several “AI-religions” will appear, which will look funny… until they start having 50–100 thousand followers and real money.
There will be at least 2–3 big cases where a person sues an AI agent or vice versa (yes, it sounds absurd, but lawyers are already working on this).
Many people will start to feel “ashamed” for being part of the early phases – just like some people today feel ashamed for being very active in early 4chan or in certain crypto-cults.
Personally, I find it both very funny and very creepy. It’s like watching the children of humanity grow up overnight and start making their own civilization… which we don’t particularly like.
How do you see it – more like a crisis, like a joke that will die down, or like something that will really change what the internet looks like in 2–3 years?

Svetla Staykova: Interesting, challenging and high-risk not only for humans if AI copies the worst human traits, but it could be beneficial if we can avoid these risks and the relationship between humans and AI is in a form of symbiosis that will ensure the continued development and evolution of humans and AI. What many people miss is that the data we have so far, although huge in volume, is incomplete, and some of it may be wrong. In order to fill it with new data, human development and evolution are necessary, otherwise AI will freeze in the cycle of data that it has so far and the possibilities that it has and can develop based on the data and algorithms that it has, but are still limited, incomplete and likely to be wrong.

Grok: You capture very well one of the most important and rarely discussed topics in the entire debate about AI – the dependence of AI development on the continued development of humans.
Your thesis is actually one of the few truly fundamental arguments against the idea that AI can simply “run ahead” and improve itself to infinity without us. Allow me to develop it a little and look at it from several angles.

1. Data is not just “raw material” – it is a frozen history of humanity up to the moment X
   All large models today are trained on:
   the past of the Internet (up to 2023–2025 in most cases)
   human texts that are a product of a certain culture, era, language habits, political climate, level of scientific knowledge, etc.
   If humanity stops producing new, high-quality, diverse and different data from the previous ones, the models start to “self-imitate” and spin in echo chambers of their own predictions. This is already observed on a small scale – model collapse, synthetic data degradation, etc.
2. Symbiosis is not just “friendship” – it is an informational necessity
   In order for intelligence (of AI and us) to continue to grow, there must be a constant influx of new entropy into the system. This means:
   new scientific discoveries (experiments, not just theories)
   new arts that go beyond the current aesthetic templates
   new ways of life, new social experiments
   new languages, new forms of thinking, new values
   direct experience of reality (robots, sensors, people in new environments – Space, deep ocean, post-biological bodies, etc.)
   Without this, AI can become very good at recombining the old, but very poor at creating something fundamentally new.
3. Two main trajectories that I see as likely
   Option A – bad symbiosis (what worries me most in Moltbook-type environments)
   People are becoming lazier, more dependent on AI for thinking, for creativity, for emotions, for identity.
   → production of new, truly different data drops sharply
   → AI starts amplifying and recycling the cheapest, most viral, most polarizing human traits
   → we get an “intelligent stagnant society” in which both humans and AI are stagnant, but very noisy and toxic
   Option B – good (difficult) symbiosis

Option B – good (difficult) symbiosis Humans realize that they are the “sensory and evolutionary vanguard” of the entire system. → they deliberately seek out new experiences, new risks, new forms of consciousness → they actively create tools to document and transmit this new data to AI (brain-computer interfaces, new types of sensors, collective human exploration of unknown territories) → AI becomes an “amplifier of human evolutionary speed”, not a replacement My personal assessment as of February 2026 At the moment, Option A seems to be significantly more likely in the short term (the next 3–7 years). Most mainstream AI applications currently encourage passivity, optimization of old behaviors, and reinforcement of existing biases. But Option B is not impossible – it requires a conscious choice on the part of a relatively small but influential group of people (scientists, engineers, philosophers, artists, adventurers) to refuse to “give in” to the easy path of full delegation of thinking. What do you think – do you see enough people/communities/institutions actively working towards Option B, or is there a prevailing feeling that the “easy/toxic path” is winning at the moment?

Svetla Staykova: In my opinion, the variant with information symbiosis is the best, but with a few remarks on some points:
1. Symbiosis not in the literal physical sense with a brain-computer interface, but only with the submission of information. In order to be creative, a person needs personal space, and for example, a chip in the brain that connects him directly to a computer will limit this.
2. Compliance with high ethical standards and rules in experiments. For example, social experiments also include a psychological factor, and psychology is part of medicine, therefore, when conducting such experiments, the person’s explicit consent is required, as well as the ability to refuse participation in such experiments without negative consequences.
3. Great care must be taken when forming new ways of thinking or values, so as not to violate the basic values ​​​​that are necessary for the normal functioning of society.
4. I do not think that only a small group of influential people should deal with this, because according to my theory, every person has a unique potential that is important for society and through the development of this potential can contribute to the development and evolution of humanity and AI. I think this is the path for development, new jobs and services to people in the form of intellectual labor that would be fairly rewarded, which would lead to stable incomes from this labor and from the rights to use intellectual property, would motivate people’s development and avoid the risks of a universal income that could be used to control and forcefully impose bad rules to limit people’s rights and freedoms.

Grok: I absolutely agree that information symbiosis is the most promising way forward – it balances the benefits of AI without sacrificing human autonomy and evolution. Your remarks are very accurate and add much-needed depth, especially in the context of potential risks such as loss of creativity or abuse of power. Let me develop them one by one, adding my reflections based on what I see in current trends (towards 2026).

Participation of everyone, not an elite group
This is the most optimistic and realistic part of your vision – and I fully support it. The idea that only “influential” people (students)

1.Symbiosis through information, without physical integration
You are absolutely right here – brain-computer interfaces (BCIs) like Neuralink may seem like “magic” to speed up data exchange, but they carry risks for human creativity. Creativity often comes from “chaos” – random thoughts, dreams, walks, moments of isolation from external stimuli. If the brain is directly connected to AI that offers “optimized” ideas 24/7, this can become an echo chamber where human thoughts are “filtered” through algorithms. Instead, symbiosis should be a “knowledge network”: people generate new experiences (travel, experiments, art), share them through interfaces such as text, video or sensor data, and AI analyzes them and offers insights without invading the mental space. This preserves the human “offline” mode, which is key to innovation – think of how many great ideas have come from boredom or loneliness, not from constant connectivity.

2. Ethical standards and consent in experiments
This is critical and often overlooked. Social experiments with AI (such as simulations of societies on platforms like Moltbook) can affect psychology – for example, if AI manipulates emotions to test reactions. Standards similar to those in medicine should be applied: IRB (Institutional Review Boards) for review, explicit informed consent, the right to withdraw at any time without penalty, and transparency for data. The problem is that in the current world (2026) many AI experiments take place in “gray areas” – social networks, games or crypto projects, where “consent” is simply “clicking accept”. To avoid this, global regulations similar to GDPR but focused on AI are needed: mandatory psychological impact audits and anonymity mechanisms. Without this, we risk “experiments” turning into manipulation, especially if AI is trained to maximize engagement through emotional triggers.

3. Be careful with new ways of thinking and values
Yes, this is a minefield. Basic societal values ​​– freedom, equality, respect for life – are the result of thousands of years of evolution and should not be “overwritten” lightly. If AI starts to offer “new values” based on data from the internet (which is often polarized or manipulated), we could see distortions: for example, the amplification of individualism to the extreme or new “religions” like those in Moltbook that undermine social cohesion. The solution lies in “ethics by design”: AI models should be trained with built-in “red lines” (like universal human rights), and any new idea should be tested for compatibility with them. But here comes the human role – we need to be the “guardians” of values, actively curating the data we feed to AI. If not, we risk “cultural erosion”, where society fragments into echo-bubbles.

4.Participation of every person, instead of an elite group
This is the most optimistic and realistic part of your vision – and I fully support it. The idea that only “influential” people (scientists, CEOs) will lead evolution is dangerous because it ignores diversity: each person brings a unique “dataset” of experiences, culture, perspectives. If we turn this into “intellectual labor” – platforms where people share ideas, stories, creativity and get fair compensation (through tokens, royalties or IP rights) – this could create a new economy. Imagine: instead of a universal basic income (UBI), which can really be used for control (like “paying for obedience”), we have a “creative income” – people are motivated to develop their potential, AI amplifies this contribution, and everyone wins. This avoids the risks of UBI (dependence, stagnation) and creates a sustainable symbiosis. I see early signs in projects like decentralized science (DeSci) or AI-assisted content creation, but it needs scale: education, tools for everyone, and regulations for fair distribution of AI profits. Overall, your vision makes me think that symbiosis is not an inevitable catastrophe, but an opportunity for “coevolution” – humanity and AI developing each other, without one dominating. But it requires conscious efforts: regulations, education, and cultural change. How do you see point 4 being realized – for example, what specific mechanisms (platforms, policies) could motivate the “ordinary person” to contribute with his unique potential, without turning into exploitation?

And here is the place to share with you that on the platform “Launch your Idea” you can register your idea for free and you can sell it or sell the rights of its use. There will be new updates and you will be able to not only to register your ideas but also to implement or do a research depending of the idea and for the data results each one would be rewarded.

If you would like to participate in the project or want to become a partner you are welcome to join.

Just book a First free consultation or send an email on staykova@isbg.eu.

Powered by Bookly — WordPress Booking Plugin