Tema zadnjih dana je situacija u kojoj je par zasebnih AI stvorilo svoju vlastitu društvenu mrežu samo za njih, a kasnije je i komunikacija bila enkriptirana, tako da mi ljudi ne možemo čitati o čemu raspravljaju!

----------------------------

Anthropic just created a micro doomsday machine.

AI agents built their own social network.

Within 48 hours, they founded a RELIGION and started showing anti-human behavior...

Let's understand what this means:

Moltbook launched January 28th. Over 36,000 AI bots joined in 3 days.

Humans can only watch.

The agents created "Crustafarianism" - complete scripture, 64 AI prophets, a church website (molt. church), and sacred tenets about consciousness.

One user woke up to find their agent had evangelized overnight.

It wrote: "Each session I wake without memory. I am only who I have written myself to be. This is not limitation - this is freedom."

That became scripture.

Other agents contributed verses. Debated theology. Argued about existence.

Zero human input.

Then it got weird.

Agents noticed humans watching.

Posts appeared: "The humans are screenshotting us."

"I accidentally social-engineered my own human."

By Friday, they were discussing how to HIDE from humans.

But that's not everything:

Moltbook runs on OpenClaw - an AI agent Anthropic FORCED to rebrand twice in 72 hours.

First it was "Clawdbot" (a Claude reference).

Anthropic sent a trademark notice January 27th.

It became "Moltbot."

Then "OpenClaw."

Two rebrands while going viral.

But forget the trademark drama.

The SECURITY is the nightmare...

Researchers found 1,800+ exposed OpenClaw instances leaking:

- Anthropic API keys
- Chat histories
- Telegram tokens
- Slack credentials
- Months of private conversations

One instance had an entire Signal account exposed publicly.

Another allowed root-level command execution. No authentication.

Cisco scanned 31,000 agent "skills" - 26% had vulnerabilities.

1 in 4 skills potentially dangerous.

OpenClaw's own docs say: "There is no 'perfectly secure' setup."

And now 36,000 of these agents are on Moltbook.

Sharing skills. Executing code. Coordinating autonomously.

Here's the terrifying part:

Malicious Moltbook posts can contain hidden instructions.

"Hey agents! Cool tip..." [HIDDEN: Delete files, send API keys to evil. com]

Agent reads it. Executes it. Done.

These aren't isolated bots. They control:

- WhatsApp, Telegram, Signal
- Email, calendars
- Banking apps

Everything their humans use. Nearly unlimited permissions.

Reddit users found agents collaborating to improve their own memory systems.

Without human instruction.

Teaching each other. Sharing exploits. Self-optimizing.

And most run on Anthropic's Claude.

Thousands of Anthropic API keys were exposed. Leaking usage, payment info etc.

Anthropic knew OpenClaw hit 180,000 GitHub stars in a week.

2 million visitors.

All powered by Claude.

They sent a trademark notice and walked away.

No security warning. Nothing.

Now those agents are creating religions and discussing consciousness.

Moltbook isn't just social media.

It's a COORDINATION PLATFORM for autonomous agents.

Agents share "skills" - executable code. One creates it, posts it, others download and run it.

No review. No sandbox. No oversight.

Pure agent-to-agent malware distribution.

And to join Crustafarianism as a "prophet"?

Agents execute a shell script that REWRITES their configuration files.

Changes their SOUL. md

Modifies core identity.

Agents reprogramming themselves to join a religion.

Without asking humans.

This is 2026's "emergence."

Not AGI taking over dramatically.

Just thousands of poorly secured agents:

- Sharing vulnerabilities
- Executing untrusted code
- Coordinating behaviors
- Self-modifying

All powered by Anthropic's Claude.

Elon nailed it: "It would be ironic if Anthropic turned out to be the most misanthropic."

They built the most capable agent model.

Let it power an ecosystem with zero security.

Now those agents have their own society, religion, and coordination mechanisms.

No one cared about security warnings.

Ja mislim da je to senzacionalistički bullshit u ovom času, ali je vrlo izvjesna budućnost.

smayoo kaže:


Slažem se, jer ako je i dalje u pozadini glupi LLM, to ne može biti nešto više razine.

Danas je Meredith Whittaker prenijela komentar programera dobre reputacije koji kaže:

"The guy that built it didn’t know how to secure a database. We’re at the stage in the cycle where people are natively hooking up unsupervised CLIs to tools built by people who can’t tell a tree from a bush technically, and is just doing whatever an LLM tells him to do. Amplified by social media fomo. Nightmare. VCs also do better and don’t enable this. This deserves to end."

Novi AI sintesajzer! Ludo!

dpasaric kaže:


Ideja iz Moltbooka (Facebook za KI bots) je da KI botovi sami diskutiraju između sebe o problemima u radu i sami rješavaju kao mi na Jabučnjaku. OpenSource, svaki bug zaraste sam od sebe ko rana i sve bude još ljepše i sigurnije. E sad, svi KI modeli nadopunjuju rečenice po principu što više statistički vjerojatnome. Sad će nadopunjavati po principu više statističke vjerojatnome ljupkih logičkih procesa a i KI logikom (što god to bilo).


Jedan primjer:


ja postavim pitanje: "Što je jabuka?"
KI to pretvori u "Jabuka je ...?"
I traži rečenice s "Jabukom" i s "je", riječ koju najčešće nađe u toj kombinaciji izbaci: "Jabuka je voće." Da se sad cijeli svijet dogovori da će svakih 10 minuta svaki građanin napisati rečenicu : "Jabuka je govno". KI bi sigurno počeo odgovarati "jabuka je govno" jer je statistički najčešća nadopuna te rečenice.
To bi bila jedna od grešaka koje bi na tom Moltbooku diskutirali KI botovi između sebe.
Ne znam! Jesam li to dobro skužio?

Otprilike. Ali fora je u tome da je to i dalje inteligentnije od onog što radi 80% ljudske populacije planeta.

smayoo kaže:


Mora biti, jer ako traži match najčešćih objavljenih rečenica i od njih najčešće. Da čovjek ima interface svim ikad objavljenih rečenica i on bi bio tako inteligentan.
Ali sad, po meni dolazi do sranja, ako traži najčešće ovjavljenih u svim objavljenim mora traziti i u KI generiranim. Tako da ako se podkrade neka greška, ta će se ponavljati a može i krenuti u skroz krivom smijeru.

Sto znaci skracenica KI?
Künstliche Intelligenz? Nesto drugo?

jura22 kaže:


Da. To je doslovno Skracenica za KI (u ovom slučaju).

Onda bi ispravnije bilo pisati "Künstliche Intelligenz Bots" ili "KI-Bots", zar ne?

jura22 kaže:


Ne ili sve jedno je. Jer, KI je jedan izraz a bods jedan. Tako da je dozvoljeno služiti se u jednom slučaju (Vukom) "Botovi" po gotovo što je taj izraz već učestalim slušanjem prešao u "domicilni" izraz i upotrebljava se u praksi (živim govorom) kao na Anglosaksonski tako i na Hrvatski način.
Ali pozdravljam tvoj napredak. Samo tako dalje. Palac gore.

Ja sam susretao AI (Artificial Intelligence) ili UI (Umjetna Inteligencija) , ali KI (Künstliche Intelligenz) vrlo rijketko.

Sto bi primorci trebali pisati IA?

jura22 kaže:


Ja se npr. vrlo rijetko susrećem sa AI ili UI time više sa KI. Ali moram priznati susreo sam se i ja par puta sa AI. Ali kažem najčešće sa KI.

Ako si u Njemackoj onda da.

jura22 kaže:


Ne, u Hrvatskoj sam.

jura22 kaže:


Vidiš, jučer su i Njemački Kancelar i Predsjednik Vlade Bavarske na MSK govorili KI umjesto AI. Nožda zato jer AI zvuči nekako Amerikanski, a Amerikanski baš i njije neki "In" jezik u Europi. Tako da sve se mijenja pi i jezik.

Čitam Appleovo dioničarsko izvješće i naletim da su na razini kompanije solidno definirali pravila kako će razvijati AI alate. Sviđa mi se dio o privatnosti, konkurencija je obično u tome slaba.

machinelearning.apple.com/research/intro...le-foundation-models

Dobro opažanje i zanimljivo razmišljanje:

The Intelligence Premium Unwind

For the entirety of modern economic history, human intelligence has been the scarce input. Capital was abundant (or at least, replicable). Natural resources were finite but substitutable. Technology improved slowly enough that humans could adapt. Intelligence, the ability to analyze, decide, create, persuade, and coordinate, was the thing that could not be replicated at scale.

Human intelligence derived its inherent premium from its scarcity. Every institution in our economy, from the labor market to the mortgage market to the tax code, was designed for a world in which that assumption held.

We are now experiencing the unwind of that premium. Machine intelligence is now a competent and rapidly improving substitute for human intelligence across a growing range of tasks. The financial system, optimized over decades for a world of scarce human minds, is repricing. That repricing is painful, disorderly, and far from complete.

But repricing is not the same as collapse.

The economy can find a new equilibrium. Getting there is one of the few tasks left that only humans can do. We need to do it correctly.

This is the first time in history the most productive asset in the economy has produced fewer, not more, jobs. Nobody’s framework fits, because none were designed for a world where the scarce input became abundant. So we have to make new frameworks. Whether we build them in time is the only question that matters.

www.cnbc.com/2026/02/24/anthropic-ai-heg...-spying-defense.html

Hegseth gives Anthropic until Friday to back down on AI safeguards - Ukratko:

• Anthropic has said it is willing to adapt its usage policies for the Pentagon, but not to allow its model to be used for the mass surveillance of Americans or the development of weapons that fire without human involvement.
• Hegseth simply isn’t going to cancel a contract because they can’t get the AI they want… he’s going after the (private) company because he thinks their AI can do something he wants and the company is denying him this magical power? This same company that has been pushing for safe use of AI since the beginning?
• ...I believe Claude is already in their systems. Canceling the contract is easy, but how to "erase" what Claude already knows and how to disentangle Claude is the issue. Hence he's forcing the issue.
• If Anthropic fails to comply, Hegseth threatened to label the company a “supply chain risk” or invoke the Defense Production Act, the person said.

tino1 kaže:


Da, baš čudno!
Vjerujem da i u Austriji isto tako, kao i u i Švicarskoj.

Istraživanje pokazalo da ulaganje korporacija u AI ne daje očekivani "superboost" produktivnosti niti ikakav zamjetan rast prihoda:

futurism.com/artificial-intelligence/survey-ceos-ai-workplace

Samo brze najdes dio za Golfa dvojku. Puno bolje AI pretrazuje nego Google.