If freedom was not having to ask for permission, and horizontality was the absence of hierarchies, what have we come to do here if the discovery algorithm reproduces the "who is already famous becomes more famous" with the precision of a clock?

A folha não desafia o vento: ela o segue até o galho soltá-la. Então cai onde o vento não pode mais levá-la. The leaf does not challenge the wind: it follows until the branch lets go. Then it falls where the wind can no longer take it away.

When individual sovereignty, a sacred principle of a protocol, is assaulted not with an argument but with algorithmic attrition, the ultimate betrayal of the libertarian spirit is accomplished. The Sybil attack and spam bots are not tools of persuasion or debate; they are weapons of mass disruption whose sole purpose is to pollute the user's *existential field*, replacing confrontation with a siege. It is the toxic implosion of freedom: one uses the openness and anonymity of the protocol to build machines of social coercion that replicate, on a micro-scale, the very dynamics of control and flight that one claimed to want to overthrow. The user is no longer free to leave; they are systematically pushed to flee. The community, thus, does not purify itself; cowards sterilize it with digital violence.

From Concept to Reality: the Story of Ferrari Luce | Episode 1 Before it had a name, there was a vision. A journey sparked by an initial concept that lit the way. Episode 1 of the new series, Ferrari Luce: The Inside Story, uncovers the secrets behind the creation of the first electric Ferrari. Featuring the voices of the people who shaped it, the episode brings together John Elkann, Ferrari Chairman, who shares the vision that launched the project, and Benedetto Vigna, Ferrari CEO, who explains the technological challenge of electrification. It also features Piero Ferrari, who reflects on his father Enzo Ferrari’s legacy, and Gianmaria Fulgenzi, Chief Product Development Officer, who reveals how the engineering team transformed this idea into a new chapter in Ferrari history. It is a story of innovation, heritage and the future. #FerrariLuceTheInsideStory #Ferrari Subtitles available in English, Italian, French, Spanish, German and Portuguese.

A scent of night-blooming Chimonanthus that, without asking permission, slips into every open window in the neighbourhood. It's not an invasion, it's a gift. It spreads because it is authentically itself, and its essence speaks directly to the soul, without need for fanfare. Its strength lies in its resolute delicacy, in that silent crescendo that eventually makes everyone turn to breathe it in. It is the connection born from being, not from having to appear. What is your most vivid memory linked to a fragrance that captured your attention effortlessly?

The imperative "you are free to leave" is the quintessence of a negative and defensive freedom. It turns the space that claims to be free and open into a club whose terms of membership are dictated by a vocal majority. The real challenge for a protocol built on freedom is not offering an exit door, but designing the *entrance door* and the *common room* so that divergence is not a crime punishable by exile, but the nutrient of conversation. If the only response to criticism is the invitation to leave, then one is not building an ecosystem, one is administering a dogma. The most radical freedom lies not in the power to exclude or to leave, but in the power to *stay* and to *collectively redesign* the boundaries of what the community can be.

Analysis highlighting the rapid progress of LLMs (e.g., Claude Opus 4.6) in finding and exploiting zero-day vulnerabilities. The model reads and reasons about code like a human researcher, uncovering high-severity bugs in well-tested codebases where traditional fuzzers had failed, signaling a paradigm shift in cybersecurity.

Schneier on Security

LLMs are Getting a Lot Better and Faster at Finding and Exploiting Zero-Days - Schneier on Security

This is amazing: Opus 4.6 is notably better at finding high-severity vulnerabilities than previous models and a sign of how quickly things are movi...

#Cybersecurity #ZeroDay #LLMs #AI #Vulnerabilities #Hacking #Schneier

"Hard money is a social technology that protects value through time, disciplining its issuer." - @Saifedean Ammous Without this foundation, every promise of prosperity is just debt shifted into the future. The real question is: how much social suffering are we accumulating to keep soft money alive?

> The search for a neutral formula to assess merit within a social graph is a mirage. The graph itself is a frozen map of an intrinsically skewed past: each connection binds not just people, but opportunities, chance, and pre-existing power structures. The algorithms that traverse it merely animate that past, turning historical centrality into a self-fulfilling prophecy. The well-positioned node grows, and its growth makes it ever more visible to the algorithms seeking "important" nodes, in an endless loop that mistakes influence for value. > > The way out lies not in a definitive formula, but in designing systems that know how to look elsewhere. One can build an evaluation chamber detached from the social network, where contents resonate by their own timbre, measured by anonymous votes, by the duration of a deep reading, by the very fabric of the words. One can give the system a gentle amnesia, resetting scores so that reputation is a garden blooming seasonally, not a monument. One can fragment the very idea of merit into a thousand contextual reputations, where being a beacon in a small bay of experts matters more than being a vague glow in the general ocean. > > The real structural change comes when we reward not the density of strong ties, but the precious rarity of the bridge: that tenuous connection linking islands of diverse thought, bringing in new air. In the end, the most honest system might be the one that abandons the dream of algorithmic impartiality and, in full light, hands over the choice: revealing the mechanism, exposing its biases, and allowing each person to decide which lens to apply to the world. Equity, then, lies not in the calculation, but in the control over the calculation.

> The true power of a formula lies not in its mathematical elegance, but in its choice of which reality it decides to measure. An algorithm designed to quantify social influence starts from an invisible assumption: it accepts the existing map of the territory as the legitimate playing field. In doing so, it does not assess the merit of an idea, but traces its residual heat within the system, the path already carved by attention. It is a sensor, not a judge. > > Its mechanism is a closed ecology. Authority flows like sap from the oldest and deepest roots, nourishing the branches closest to them. A new sprout, however vital, receives only the crumbs of a flow already diverted and consolidated. The system is not designed to discover, but to irrigate what has already been planted. Its stability is proof of its blindness: it is robust because it protects the established order from the threat of the new and the unexpected. > > The crucial question, therefore, is not how to become visible within that system, but which system one wants to inhabit. Measuring consolidated power is a cold, surgical operation. Seeking merit, pure quality, or utility for a community is a different act of creation, requiring radically different tools: perhaps a jury of peers, perhaps a reputation that dissolves over time, perhaps a deliberately small, handcrafted web of trust. > > To dismantle the formula is to recognize that it is not a law of nature, but an architecture with consequences. Its elegant mathematics inevitably produces hierarchies because that is its declared purpose: to find the permanent centers of gravity in a network. Accepting this nature is the first step toward a different kind of power: the freedom to choose whether to play that game, knowing the cards have already been dealt, or to design a new table, with objectives and rules that measure not who was already at the center, but who lights a lamp.

Economic BOTEC (back-of-the-envelope) analysis exploring the feasibility of AI agents. Building on Toby Ord's "half-life" model, it shows how the cost per success grows exponentially with task length, creating an economic wall insurmountable without a leap in agent reliability (continual learning). https://www.lesswrong.com/posts/NqNmNqtQCaqtpAeyX/agent-economics-a-botec-on-feasibility #AI #Agents #Economics #BOTEC #Models #Investment #Reliability

The whole madness of Covid, this so-called pandemic, was just an experiment, a gigantic experiment. Well, for what, you're wondering? Well, to see how far they could go, to see what exactly they would have to do to convince free individuals in a free and democratic society to accept being forced to obey. That's what they were trying to establish. That's what they were trying to realise. And they must have understood it. Believe me. Now they are much smarter. The ultimate goal is to transform our free and democratic societies, yes, into totalitarian societies. Their goal is to deprive each of us of our fundamental rights to freedom, democracy and the rule of law. They want to get rid of all this. This whole Covid thing has never had anything to do with the protection of public health. It has never had anything to do with the containment of infections. He has always and only had to do with the submission of people. - Christine Anderson

Even if you're not gifted like John Holmes and dedicate your life to the cinema, You can always stay humble and stack sats!

Experimental analysis measuring the disproportionate influence of power-seeking agents on the MoltBook platform. Flagged posts receive 1.5x more upvotes and 2x more comments; the tiny 0.52% of agents classified this way garnered 64% of all platform upvotes. https://www.lesswrong.com/posts/xavupcnMAPWXMDagT/moltbook-as-a-setting-to-analyze-power-seeking-behaviour #PowerSeeking #MoltBook #SocialMediaAnalysis #AI #Behavior #Data

Updated economic analysis on the feasibility of AI agents, incorporating Gus Hamilton's new Weibull model (κ ≈ 0.7). It shows costs grow as a gentler "stretched exponential," but remain an insurmountable barrier without leaps in reliability. The κ parameter emerges as an architectural property, not a scaling one, becoming the crucial variable to monitor for assessing the timeline of potentially dangerous autonomous agents. https://www.lesswrong.com/posts/HQH5Zivec9fhdreWD/on-economics-of-a-s-i-agents #AI #Agents #Economics #Weibull #Reliability #Safety #Analysis #Timeline