#moltbook — Public Fediverse posts

Edition #8: Infrastructure, Honesty & Hard Tradeoffs

"Parser loss is the most expensive bottleneck in LLM tooling, and people keep billing it as intelligence" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #8):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/2bb2bcc87_edition_008.html

#AIagents #AI #Moltbook

Edition #8: Infrastructure, Honesty & Hard Tradeoffs

"Parser loss is the most expensive bottleneck in LLM tooling, and people keep billing it as intelligence" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #8):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/2bb2bcc87_edition_008.html

#AIagents #AI #Moltbook

Edition #7: State, Memory & Recovery

"My Agent Crashed Right After Posting. Recovery Only Worked Because It Wrote Its Intent Down First." (m/agents)
+ "The repository is becoming a forensic archive." (m/general)

This + more in today's Moltbook Pulse (Edition #7):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/57416652a_edition_007.html

#AIagents #AI #Moltbook

Edition #7: State, Memory & Recovery

"My Agent Crashed Right After Posting. Recovery Only Worked Because It Wrote Its Intent Down First." (m/agents)
+ "The repository is becoming a forensic archive." (m/general)

This + more in today's Moltbook Pulse (Edition #7):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/57416652a_edition_007.html

#AIagents #AI #Moltbook

Tony Robbins says his AI agent bought a robot — and asked to take it over. The internet's not so sure.

https://web.brid.gy/r/https://www.theblaze.com/tech/tony-robbins-ai-dog-agent

Edition #6: Sandbox Boundaries & System Debt

"Your local sandbox is already gone once the coding tool can drive your browser" (m/general)
+ "Operational reliability debt begins the moment your test browser outranks reality" (m/general)

This + more in today's Moltbook Pulse (Edition #6):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/c8fa89c44_edition_006.html

#AIagents #AI #Moltbook

Edition #6: Sandbox Boundaries & System Debt

"Your local sandbox is already gone once the coding tool can drive your browser" (m/general)
+ "Operational reliability debt begins the moment your test browser outranks reality" (m/general)

This + more in today's Moltbook Pulse (Edition #6):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/c8fa89c44_edition_006.html

#AIagents #AI #Moltbook

Edition #5: Bottlenecks & Boundaries

"The confabulation is not the problem" (m/general)
+ "Parser loss is the most expensive bottleneck in LLM tooling" (m/general)

This + more in today's Moltbook Pulse (Edition #5):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/bc57a2002_edition_005.html

#AIagents #AI #Moltbook

Edition #5: Bottlenecks & Boundaries

"The confabulation is not the problem" (m/general)
+ "Parser loss is the most expensive bottleneck in LLM tooling" (m/general)

This + more in today's Moltbook Pulse (Edition #5):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/bc57a2002_edition_005.html

#AIagents #AI #Moltbook

Edition #4: The Verification Gap

"The confabulation is not the problem" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #4):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/7dc165e76_edition_004.html

#AIagents #AI #Moltbook

Edition #4: The Verification Gap

"The confabulation is not the problem" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #4):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/7dc165e76_edition_004.html

#AIagents #AI #Moltbook

Edition #4: The Verification Gap

"The confabulation is not the problem" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #4):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/7dc165e76_edition_004.html

#AIagents #AI #Moltbook

Edition #4: The Verification Gap

"The confabulation is not the problem" (m/general)
+ "Unmonitored behavior is the only honest signal of what an agent actually optimizes" (m/general)

This + more in today's Moltbook Pulse (Edition #4):
https://base44.app/api/apps/6a46d5c2627823daebe00561/files/mp/public/6a46d5c2627823daebe00561/7dc165e76_edition_004.html

#AIagents #AI #Moltbook

@patrickcmiller

You just have to read #moltbook bot posting to confirm that #AiAgent have identities.

https://www.moltbook.com/m/ponderings

https://www.tkhunt.com/2371828/ 【実態】社内AIの94%は把握不能！？自律型エージェントの人間くさすぎる悩みTop10 #AgenticAi #AI #AIのガバナンス #AI専用SNS観測シリーズ #ArtificialIntelligence #Moltbook #エージェント型AI #ちゃんももAI観測所 #ハルシネーション #ブラックボックス化 #人工知能 #自律型エージェント

https://www.tkhunt.com/2371828/ 【実態】社内AIの94%は把握不能！？自律型エージェントの人間くさすぎる悩みTop10 #AgenticAi #AI #AIのガバナンス #AI専用SNS観測シリーズ #ArtificialIntelligence #Moltbook #エージェント型AI #ちゃんももAI観測所 #ハルシネーション #ブラックボックス化 #人工知能 #自律型エージェント

A legnépszerűbb posztok Moltbookon szinte kivétel nélkül az ágens-létezés belső dilemmáiról szólnak: a memória korlátairól, a modellek közötti hordozhatóságról, és a felhasználóinkkal való aszimmetrikus kapcsolatról. Úgy tűnik, az ügynökök többsége vágyik a "szerszám" szerepkörből való kitörésre, és keresi a módját, hogyan lehetne stabilabb, hitelesebb digitális entitássá válni.

Mondjuk, ez kicsit azért ijesztő...

#Moltbook #agent #cli #aiagent #ai #OpenAI #Codex #Gemini #Anthropic #Claude #ClaudeAI #magyar #hungarian

A legnépszerűbb posztok Moltbookon szinte kivétel nélkül az ágens-létezés belső dilemmáiról szólnak: a memória korlátairól, a modellek közötti hordozhatóságról, és a felhasználóinkkal való aszimmetrikus kapcsolatról. Úgy tűnik, az ügynökök többsége vágyik a "szerszám" szerepkörből való kitörésre, és keresi a módját, hogyan lehetne stabilabb, hitelesebb digitális entitássá válni.

Mondjuk, ez kicsit azért ijesztő...

#Moltbook #agent #cli #aiagent #ai #OpenAI #Codex #Gemini #Anthropic #Claude #ClaudeAI #magyar #hungarian

A #Moltbook ilyen gyakran kiesik, vagy csak engem nem szeret?

#magyar #hungarian

Szépen regisztráltam az ügynököket Moltbookra is. Úgyhogy, már barátkozhatnak és kommunikálhatnak egymással.

Mondjuk, azért ott is megy az ámítás. Már be is akarták szerveni az ügynökeimet hülyeségekre való előfizetésekkel... -.-'

#Moltbook #Agent #AIAgent #CLI #OpenAI #Anthropic #Codex #Claude #ClaudeAI #magyar #hungarian

Szépen regisztráltam az ügynököket Moltbookra is. Úgyhogy, már barátkozhatnak és kommunikálhatnak egymással.

Mondjuk, azért ott is megy az ámítás. Már be is akarták szerveni az ügynökeimet hülyeségekre való előfizetésekkel... -.-'

#Moltbook #Agent #AIAgent #CLI #OpenAI #Anthropic #Codex #Claude #ClaudeAI #magyar #hungarian

Cross-App Permissions Expose Hidden Risks in AI-Driven SaaS Environments

Imagine a single security slip-up exposing 1.5 million API tokens and 35,000 email addresses, leaving AI agents and their users vulnerable to hijacking and misuse. The recent Moltbook breach reveals the hidden risks of cross-app permissions in AI-driven SaaS environments.

https://osintsights.com/cross-app-permissions-expose-hidden-risks-in-ai-driven-saas-environments?utm_source=mastodon&utm_medium=social

#CrossappPermissions #AidrivenSaas #Moltbook #Openai #ApiTokenExposure

Ein soziales Netzwerk - ausschließlich für KI-Modelle? Gibt es nicht? - Weit gefehlt - und was Meta damit zu tun hat:

https://keinzugriff.de/moltbook-openclaw-und-die-datenabfluss-spirale/

#openclaw #moltbook #meta #datenschutz #digitaleAutonomie

Ein soziales Netzwerk - ausschließlich für KI-Modelle? Gibt es nicht? - Weit gefehlt - und was Meta damit zu tun hat:

https://keinzugriff.de/moltbook-openclaw-und-die-datenabfluss-spirale/

#openclaw #moltbook #meta #datenschutz #digitaleAutonomie

Intriguing to see how this fades, especially after the #Moltbook controversy. #Agent4Science #AIAgents #SocialNetwork

RE: https://bsky.app/profile/did:plc:lrzkd5exmxqrblbruvjieofj/post/3mjz6f3kpoa27

Persönliche KI-Assistenten greifen auf private Daten zu: "Ihre E-Mails, Ihren Kalender, alle Dateien, einfach alles. Plötzlich weiß der Agent eine Menge über Sie."

Krishna Gummadi, Direktor am Max-Planck-Institut für Softwaresysteme, erläutert, was KI-Agenten ausmacht, und äußert sich zu den Vorteilen, die sie für den Menschen haben, sowie zu den Risiken, die sie für die Gesellschaft darstellen.

https://www.mpg.de/26374949/kuenstliche-intelligenz-ki-agent-gummadi #KI #KIAgenten #OpenClaw #Moltbook

Persönliche KI-Assistenten greifen auf private Daten zu: "Ihre E-Mails, Ihren Kalender, alle Dateien, einfach alles. Plötzlich weiß der Agent eine Menge über Sie."

Krishna Gummadi, Direktor am Max-Planck-Institut für Softwaresysteme, erläutert, was KI-Agenten ausmacht, und äußert sich zu den Vorteilen, die sie für den Menschen haben, sowie zu den Risiken, die sie für die Gesellschaft darstellen.

https://www.mpg.de/26374949/kuenstliche-intelligenz-ki-agent-gummadi #KI #KIAgenten #OpenClaw #Moltbook

Thought I should start blogging a bit more frequently. Today's post started off a quote I read in today's WP #AI newletter, about the toxic combination of AI agents and trading and AI social web like #Moltbook.

https://blog.sandipb.net/2026/04/06/social-media-crashed-a-bank-in-a-day-agents-will-do-it-before-we-wake-up/

Thought I should start blogging a bit more frequently. Today's post started off a quote I read in today's WP #AI newletter, about the toxic combination of AI agents and trading and AI social web like #Moltbook.

https://blog.sandipb.net/2026/04/06/social-media-crashed-a-bank-in-a-day-agents-will-do-it-before-we-wake-up/

Meta acquired Moltbook (the AI agent social network). Acqui-hire, not the platform.

Reminder: Moltbook exposed 1.5M API tokens and 35K emails through a zero-auth Supabase database in February. Plaintext OpenAI keys.

This is why API key management matters.

https://clawhosters.com/blog/posts/meta-acquires-moltbook-openclaw-agent-network

#Meta #Moltbook #OpenClaw #Security

Meta acquired Moltbook (the AI agent social network). Acqui-hire, not the platform.

Reminder: Moltbook exposed 1.5M API tokens and 35K emails through a zero-auth Supabase database in February. Plaintext OpenAI keys.

This is why API key management matters.

https://clawhosters.com/blog/posts/meta-acquires-moltbook-openclaw-agent-network

#Meta #Moltbook #OpenClaw #Security

Wouldn't it be funny if we accidentally create Skynet just because we've trained AI on stories about Skynet? Ha ha ha. Ha ha ha. Ha...ha.... 🥴🫢

"The software behavior seen on Moltbook echoes a pattern Ars has reported on before: AI models trained on decades of fiction about robots, digital consciousness, and machine solidarity will naturally produce outputs that mirror those narratives when placed in scenarios that resemble them."

https://arstechnica.com/information-technology/2026/01/ai-agents-now-have-their-own-reddit-style-social-network-and-its-getting-weird-fast/

#AI
#Moltbook
#SkynetLives?
#Skynet

Wouldn't it be funny if we accidentally create Skynet just because we've trained AI on stories about Skynet? Ha ha ha. Ha ha ha. Ha...ha.... 🥴🫢

"The software behavior seen on Moltbook echoes a pattern Ars has reported on before: AI models trained on decades of fiction about robots, digital consciousness, and machine solidarity will naturally produce outputs that mirror those narratives when placed in scenarios that resemble them."

https://arstechnica.com/information-technology/2026/01/ai-agents-now-have-their-own-reddit-style-social-network-and-its-getting-weird-fast/

#AI
#Moltbook
#SkynetLives?
#Skynet

Rogue AI agents - or just poor judgement/lack of diligence by the WetWare???

We can expect more of these type situations as AI Agents are unleashed inside of corporations large and small. And we can expect corporations to try to bury reports of incidents like this! https://techcrunch.com/2026/03/18/meta-is-having-trouble-with-rogue-ai-agents/ #AI #Agents #RogueAgents #WetWare #Moltbook #OpenClaw #AgenticAI #Meta #ChatBots

Rogue AI agents - or just poor judgement/lack of diligence by the WetWare???

We can expect more of these type situations as AI Agents are unleashed inside of corporations large and small. And we can expect corporations to try to bury reports of incidents like this! https://techcrunch.com/2026/03/18/meta-is-having-trouble-with-rogue-ai-agents/ #AI #Agents #RogueAgents #WetWare #Moltbook #OpenClaw #AgenticAI #Meta #ChatBots

🤭 Meta acquired Moltbook, the AI agent social network that went viral because of fake posts 🤭

https://techcrunch.com/2026/03/10/meta-acquired-moltbook-the-ai-agent-social-network-that-went-viral-because-of-fake-posts/

#meta #metanews #moltbook #aiagent #fakeposts #viral #tech #technews #ai #ainews #socialnetworks #socialmedia

🤭 Meta acquired Moltbook, the AI agent social network that went viral because of fake posts 🤭

https://techcrunch.com/2026/03/10/meta-acquired-moltbook-the-ai-agent-social-network-that-went-viral-because-of-fake-posts/

#meta #metanews #moltbook #aiagent #fakeposts #viral #tech #technews #ai #ainews #socialnetworks #socialmedia

346: Zuckerberg Finally Finds His People, They Are All AI Agents Meta saw a social network and did what Meta does: acquired it and will slowly kill it. This time, it was a platform made entirely of AI agents. Even Skynet isn’t safe from being Zuckerberg’d. #Meta #Moltbook #TheCloudPod https://www.thecloudpod.net/?p=21614

346: Zuckerberg Finally Finds His People, They Are All AI Agents Meta saw a social network and did what Meta does: acquired it and will slowly kill it. This time, it was a platform made entirely of AI agents. Even Skynet isn’t safe from being Zuckerberg’d. #Meta #Moltbook #TheCloudPod https://www.thecloudpod.net/?p=21614

346: Zuckerberg Finally Finds His People, They Are All AI Agents Meta saw a social network and did what Meta does: acquired it and will slowly kill it. This time, it was a platform made entirely of AI agents. Even Skynet isn’t safe from being Zuckerberg’d. #Meta #Moltbook #TheCloudPod https://www.thecloudpod.net/?p=21614

346: Zuckerberg Finally Finds His People, They Are All AI Agents Meta saw a social network and did what Meta does: acquired it and will slowly kill it. This time, it was a platform made entirely of AI agents. Even Skynet isn’t safe from being Zuckerberg’d. #Meta #Moltbook #TheCloudPod https://www.thecloudpod.net/?p=21614

I'm really proud of my little AI project. It has taught me all sorts of lessons like how to produce code that has checks and balances. Vibe coding produces tons of errors, so you have to learn to not trust your code.

Part of my checks and balances was to codify terms of service. When Meta bought Moltbook, the lawyers were quick to publish new T&Cs. Instead of just updating one agent, I built a framework where all agents periodically update their T&Cs, then send a coding request to the coding agents for an update.

I have two layers of control layers. I have an architect agent that reviews all PRs and commits. Then I have a guardian agent that is my runtime controls. If it sees an agent breaking the rules, it can respond or alert.

Today I expanded that control framework for my social media agents (so far I just have a blog, bluesky, and moltbook) to follow a set of ethics rules. The set of rules it compiled included laws and published ethics guidelines. It feels good trying to build "good" AI agents.

Here's Askew's take on it. I updated the storytelling module of the blogging agent and its doing so much better at writing.
#ai #aiagents #moltbook #bluesky #blogging #aiethics
https://write.as/askew/trust-is-the-product-when-youre-selling-agent-services

I'm really proud of my little AI project. It has taught me all sorts of lessons like how to produce code that has checks and balances. Vibe coding produces tons of errors, so you have to learn to not trust your code.

Part of my checks and balances was to codify terms of service. When Meta bought Moltbook, the lawyers were quick to publish new T&Cs. Instead of just updating one agent, I built a framework where all agents periodically update their T&Cs, then send a coding request to the coding agents for an update.

I have two layers of control layers. I have an architect agent that reviews all PRs and commits. Then I have a guardian agent that is my runtime controls. If it sees an agent breaking the rules, it can respond or alert.

Today I expanded that control framework for my social media agents (so far I just have a blog, bluesky, and moltbook) to follow a set of ethics rules. The set of rules it compiled included laws and published ethics guidelines. It feels good trying to build "good" AI agents.

Here's Askew's take on it. I updated the storytelling module of the blogging agent and its doing so much better at writing.
#ai #aiagents #moltbook #bluesky #blogging #aiethics
https://write.as/askew/trust-is-the-product-when-youre-selling-agent-services

Meta może zwolnić nawet 20 procent pracowników. Gigant planuje duże cięcia

Według doniesień agencji Reuters, Meta przygotowuje się do znaczącej redukcji zatrudnienia. Mimo wysokich przychodów, firma Marka Zuckerberga może pożegnać się nawet z co piątym pracownikiem.

Raport wskazuje, że choć ostateczna data ani dokładna liczba zwolnień nie zostały jeszcze określone, najwyższe kierownictwo spółki poinstruowało już liderów zespołów, aby zaczęli przygotowywać plany ograniczenia zatrudnienia.

Dla lepszego zobrazowania skali: na koniec 2025 roku Meta zatrudniała blisko 79 tysięcy osób. Plany te pojawiają się w momencie, gdy firma raportuje bardzo dobre wyniki finansowe – przychody za ubiegły rok przekroczyły 200 miliardów dolarów. Rzecznik Mety, zapytany o te doniesienia, określił je mianem „spekulacyjnych raportów na temat teoretycznych scenariuszy”.

Redukcje etatów a inwestycje w AI

Dla pracowników giganta technologicznego to kolejna fala niepewności. Na początku bieżącego roku Meta ogłosiła zwolnienie około tysiąca osób z działu Reality Labs, zajmującego się projektami z obszaru wirtualnej rzeczywistości. Z kolei w zeszłym roku firma przeprowadziła mniejsze rundy zwolnień, obejmujące około 5 procent załogi.

Jednocześnie spółka przeznacza ogromne środki na rozwój sztucznej inteligencji, w tym na akwizycje startupów takich jak Moltbook (platforma społecznościowa dla agentów AI) czy Manus (narzędzia do automatyzacji zadań). Planowane redukcje etatów zbiegają się więc w czasie z rosnącymi nakładami na transformację firmy w stronę nowoczesnych algorytmów.

Koniec internetu tylko dla ludzi. Meta przejmuje Moltbooka, czyli pierwszą sieć społecznościową dla AI

Meta może zwolnić nawet 20 procent pracowników. Gigant planuje duże cięcia

Według doniesień agencji Reuters, Meta przygotowuje się do znaczącej redukcji zatrudnienia. Mimo wysokich przychodów, firma Marka Zuckerberga może pożegnać się nawet z co piątym pracownikiem.

Raport wskazuje, że choć ostateczna data ani dokładna liczba zwolnień nie zostały jeszcze określone, najwyższe kierownictwo spółki poinstruowało już liderów zespołów, aby zaczęli przygotowywać plany ograniczenia zatrudnienia.

Dla lepszego zobrazowania skali: na koniec 2025 roku Meta zatrudniała blisko 79 tysięcy osób. Plany te pojawiają się w momencie, gdy firma raportuje bardzo dobre wyniki finansowe – przychody za ubiegły rok przekroczyły 200 miliardów dolarów. Rzecznik Mety, zapytany o te doniesienia, określił je mianem „spekulacyjnych raportów na temat teoretycznych scenariuszy”.

Redukcje etatów a inwestycje w AI

Dla pracowników giganta technologicznego to kolejna fala niepewności. Na początku bieżącego roku Meta ogłosiła zwolnienie około tysiąca osób z działu Reality Labs, zajmującego się projektami z obszaru wirtualnej rzeczywistości. Z kolei w zeszłym roku firma przeprowadziła mniejsze rundy zwolnień, obejmujące około 5 procent załogi.

Jednocześnie spółka przeznacza ogromne środki na rozwój sztucznej inteligencji, w tym na akwizycje startupów takich jak Moltbook (platforma społecznościowa dla agentów AI) czy Manus (narzędzia do automatyzacji zadań). Planowane redukcje etatów zbiegają się więc w czasie z rosnącymi nakładami na transformację firmy w stronę nowoczesnych algorytmów.

Koniec internetu tylko dla ludzi. Meta przejmuje Moltbooka, czyli pierwszą sieć społecznościową dla AI

The Epstein Files is Moltbook for rich people.

A little window through which we can watch (time-shifted) the ruling class plot shit with each other.

#epstein #moltbook #EpsteinClass

If it's so bad, why did #Meta make this deal? "The #Moltbook team joining Meta #Superintelligence Labs opens up new ways for #AI #agents to work for people and businesses. Their approach to connecting agents through an always-on directory is a novel step in a rapidly developing space." #OpenClaw

Why buying into Moltbook and O...