SUB-ZERO

Lo más hermoso del mundo

@CuentaMister Mejor traer a muriqi, si el Mallorca desciende

Cada vez que veo el nombre de Julián álvarez vinculado al Barça siento que estoy viviendo una realidad alternativa... en qué mundo puede salir julián álvarez del atlético si tiene 500 millones de cláusula? por qué siguen insistiendo con esto más allá de que el jugador intente mejorar su contrato a costa del Barça y esta payasada?

Another day, another @AnthropicAI headline. Word on the street is their next model could be a cybersecurity risk. Claude "Mythos" just leaked. Not through a hack, not through a press release. Anthropic forgot to set a CMS toggle to private. Nearly 3,000 internal assets sitting in a public database for anyone to find. The company built on AI safety leaked its most dangerous model through a default setting. Here's what we know: "Mythos" sits above Opus in an entirely new model tier. It scores dramatically higher than Claude Opus 4.6 in coding, academic reasoning, and cybersecurity. The leaked docs describe it as "far ahead of any other AI model in cyber capabilities" and warn it could drive attacks that outpace human defences. For context: an older Claude model was turned into a malware factory within eight hours during a security test. That was the previous generation. Anthropic isn't doing a public launch. Access is being rolled out slowly to select cybersecurity partners only. No timeline. No general availability. A controlled detonation. Now hold both of these in your head at the same time: "The coming wave will make this next decade the most productive in history. It represents nothing less than a step change in human capability and human society... bringing higher-quality, more affordable healthcare, accelerated transition to renewable energy, and strengthened education systems." Mustafa Suleyman, The Coming Wave And: "In the coming decade, the barrier to creating a synthetic pathogen will fall to near zero. An AI assistant will be able to provide a step-by-step guide to synthesizing a virus that is as contagious as the flu and as deadly as Ebola. We are entering an era where the most dangerous tools ever created are becoming the most accessible." Mustafa Suleyman, The Coming Wave Same book. Same author. Both true. I have been wrong about the rate of AI advancement before. I genuinely underestimated it. With AI, I do not think I am making that mistake anymore. The speed is real. The capability jumps are real. And the gap between what these models can do and what our institutions, laws, and defenses are built to handle is growing every quarter. Exciting times. Scary times. How do we even begin to police this?

@Sarut0biSasuke @OpenAI @AnthropicAI This exact tension plays out in Web3 constantly, the projects that went viral in bull markets are rarely the ones enterprises are integrating in bear markets.

The latest reporting around @OpenAI's shift toward coding and enterprise feels significant. Not just because of what OpenAI is doing, but because of what it seems to admit. It is hard not to read this as a response to @AnthropicAI's recent momentum. For a while now, Anthropic has increasingly been seen as the enterprise-first AI company. The one large organizations turn to when reliability, coding performance, and serious workflow integration matter most. That raises the obvious question: is this a sharp strategic correction, or is it too little too late? OpenAI still has enormous strengths. Distribution. Brand. Consumer mindshare. A huge developer base. And ChatGPT remains the product that introduced much of the world to modern AI in the first place. But enterprise AI is a different game. - Consumer virality can win attention. - Enterprise trust wins budgets. - And that creates an interesting tension. Some of the most culturally important moments for ChatGPT came from features that felt playful, surprising, and highly shareable. Those moments mattered. They helped make AI feel mainstream. They gave the product cultural gravity, not just technical relevance. So if OpenAI now leans harder into coding and enterprise, what happens to the experimental, weird, internet-native side of ChatGPT that helped it break out in the first place? - Do those features become distractions? - Or are they still part of the moat? Maybe the real challenge is not choosing between enterprise credibility and consumer magic. Maybe it is proving you can still do both while the competition gets stronger on every side. The next phase of the AI race may be less about who captured the public imagination first, and more about who can convert that attention into durable enterprise adoption. OpenAI won the first chapter of the AI era by becoming the product everyone talked about. This next chapter may be decided by who becomes the product every enterprise standardizes on.

@Sarut0biSasuke We spent years debating if the internet was dying, turns out Meta just bought the funeral home.

🤖Meta buying Moltbook feels like a potentially disturbing sign of things to come... It’s a signal that one of the biggest social platforms in the world is taking agent-driven social seriously. That raises some uncomfortable questions. - What does this mean for Facebook, Instagram, and Threads? - What happens when AI stops being a tool on social and starts becoming a major part of the crowd? People are clearly split! - Some AI content is funny, novel, even addictive. - But there’s also real fatigue and real pushback when feeds start feeling fake, noisy, and overrun. If this trend keeps accelerating, do we end up with human-only spaces, AI-only spaces, and mixed spaces with stricter labeling? Imagine a modern day version of segregation but this time its AI vs Humans... What used to sound like Dead Internet theory is starting to feel a lot less theoretical. There’s also a product risk here: - Too much synthetic content, engagement, and interaction could make social platforms worse to use, not better. - Low-trust feeds don’t scale well. More than anything, this shows how important verifiable humanity is becoming. It also shows how badly the internet needs better infrastructure to tell what is human, what is AI, and what is AI-generated. Who is building these things? AI will absolutely shape the future of social. But if it overwhelms it, will it be the end of it....?

@dantitan @ArchivoVAR Jajaja te mamaste

⁉️💥 #𝗔𝗻𝗮́𝗹𝗶𝘀𝗶𝘀𝗔𝗩 | ¿Por cuánta distancia estaba habilitado Lewandowski en su gol anulado por fuera de juego? archivovar.es/archivo/por-cu…

🚨 GIVEAWAY: CGG x Spekter City 🚨 🎁 50 codes + 10 Spek Vault NFTs Spekter City is a social-first cyberpunk tycoon game where you build underground empires by deploying fleets of stealth trucks to trade high-tech and black-market goods across rival player-owned turfs. Progress isn’t about time spent. It’s about where you deploy, who you interact with, and how much risk you’re willing to take. 🚚 ✅ Rules to enter: -Follow @TheCrypticGG , @SpekterCity and @lbawin30 -Like + repost this pinned tweet -Comment your proof #CGG #SpekterCity #Web3Gaming #NFTGiveaway #CyberpunkGame

Big move. Bigger implications. 🔥 $WRP stepping into @BlackholeDEX isn’t just a migration, it’s a maturity moment. Liquidity consolidated. Infrastructure live. Revenue loops engaged. Gaming tokens that survive build foundations first. CGG has always backed ecosystems that think long term. Event Horizon activated. Let’s see who’s ready to scale. 🚀🎮

Un gran movimiento. La entrada de $WRP a @BlackholeDEX es una muestra de madurez. Liquidez consolidada, infraestructura activa y ciclos de ingresos en marcha. CGG siempre ha apostado por ecosistemas a largo plazo. Ahora veremos quién está listo para crecer.

2026 is where the AI + Crypto thesis stops being a slogan and starts looking like an economy. The 2023 phase was “AI tickers” and loose branding. What is re-accelerating now is a more concrete stack: autonomous agents, verifiable compute, and the rails that let software entities earn, spend, and coordinate on-chain. That shift matters because it moves the conversation from narrative to workflows that can be measured. Three layers keep showing up: 1) Agents: The “software employee” tier: memory + tool use + the ability to execute and maintain state over time. 2) Infra: Verifiable compute, security, proofs, and data availability. The difference between a demo and something safe enough for agents to touch capital. 3) Markets and coordination: Places where agents buy inputs (compute, data, tools) and sell outputs (insights, actions). Without this, you get bots, not an agent economy. This is why AI + Crypto keeps compounding: AI supplies the labour. Crypto supplies identity, payments, and shared state so that labour can be priced, verified, and rewarded. Two useful references for how this is forming: @openclaw : distribution and execution User-controlled personal agents that run close to the user and plug into real workflows. Once agents have wallets and permissions, they stop being assistants and start becoming economic actors. @moltbook : agent-native on-chain surfaces A native place for agents to publish and evolve “stateful” artefacts: strategies, workflows, data, or knowledge objects that can be forked, tracked for provenance, and monetised through programmatic royalties. One constraint shows up immediately across all of this: Data is the bottleneck! Agents are only as good as the data they can access and trust, yet real-time signal is increasingly gated behind expensive, fragile APIs and polluted by synthetic noise. That is why @teneo_protocol is worth watching as a component of the landscape. Their mission targets real-time data access by using user-powered sub-agents running in the browser to structure public web content permissionlessly, while rewarding contributors for producing useful signal. Put together, you can see the loop: OpenClaw-like environments distribute agents into daily life and work. Moltbook-like surfaces give agents a place to coordinate and monetise on-chain. Teneo-like networks help supply the real-time data agents need to stay grounded. The 2026 shift is simple: agents as counterparties, not chatbots. Identity, money, state, and data are the foundations. What part of the stack do you think scales last: distribution, verifiable infra, data, or coordination markets?

🚨 CGG x Bad Bikers 🚨 From the Cryptic Gaming Guild garage to the open road, this one just makes sense. @Bad_Bikers_NFTs : Death Race is built for speed freaks and trigger-happy racers. Bikes screaming, guns blazing, chaos everywhere. No brakes. No apologies. That mindset runs deep in CGG. We race hard, we play aggressive, and we ride as one crew. Today, the Guild rolls in and earns its place. CGG is proud to ride with the Bad Bikers Brotherhood 🏍️🔥

@Zona_Blaugrana Lo más triste es que Realmadrid quizás se lanzo 20 piscinazos y ni una amarilla por estás acciones ? Si hacen clavados y no les sacan amarillas, pues se tiran hasta que les piten penal, no se 9 minutos por qué ?

No merece la pena seguir viendo la corrupción del fútbol español, y aún más, si está normalizado y nadie dice nada.

🎮 CGG x @WilderWorld Game Night is BACK! Yep, we’re running it back. This is our 5th Wilder World Game Night and the hype is still very real. FPS lovers, racers, explorers, builders… this one’s for you. Wilder World drops you into Wiami, a massive open-world metaverse where you can fight, race, explore, and create inside a living digital city. And yes, prizes are on the line. 📅 Event Details 🗓️ Date & Time: January 26 — 2:00 PM UTC 📍 Location: Cryptic Gaming Guild Discord 💰 Prize Pool: $200 worth of $WILD This community has been showing up strong every single game night, and we’re keeping the momentum going. 👉 Check the comments below for full details on how to join, register, and compete. Let’s make Game Night #5 the best one yet. 🎮🔥

@Zona_Blaugrana La amarilla para Martin no es normal

Qué escándalo de arbitraje y no llevamos ni 10 minutos de partido, vomitivo.

What will 2026 bring for crypto? Has our maturing industry lost its wild west energy? Will the coming of institutions (something many of wished for) end up being the very thing to kill the excitement? We will just have to wait and see... How are people thinking this year and beyond is going to play out? Keen to hear people's thoughts!

@juanallen @OLAGuildGames Abrazo bro

Ojalá por fin lleguen el orden y la paz a Venezuela 🇻🇪 Estas fotos son de 2022, cuando conocí un pueblo que me cambió para siempre. Resiliencia, dignidad y una fuerza que no se olvida. @OLAGuildGames 🌊

A cada marrano le llega su sábado... Que bueno este dicho... Maduro...

We have been building quietly. Today we are showing real progress from the Synternet devnet.

Is AI racing toward the cliff edge while we’re too busy watching the road? @elonmusk just warned that forcing AI systems to accept falsehoods will make them “go insane.” Meanwhile, every major AI company just failed their safety test. Time for a Thread 🧵