The loudest voices stoking fears about AI dangers have made tremendous headway in the past two weeks. AI technology has not taken some unexpected, dangerous turn, but the hype around it — propelled by what appears to be a well orchestrated PR campaign — has drummed up considerable fear. I worry that it represents a setback for our field. I have written frequently that fears of AI are overhyped. AI’s capabilities can be uncannily human-like and unpredictable, and it’s rational to worry when people who are directly involved express concerns. But I see the problems as a sign of the engineering work that ahead, rather than insurmountable barriers or the sky falling. AI technology continues to advance — which is a good thing! — but technical advances, poorly understood by the public, give those who seek to generate hype repeated opportunities to do so. First, I don’t see any step up in the risk of human extinction from AI compared to a few months ago. The theories about this remain the same fantastical, science fiction scenarios as a few months ago. The biggest change in AI risk is its cybersecurity capabilities — a topic which we should take seriously — but this, too, will not lead to the end of the world. The most notable recent event leading to increased fear was when an OpenAI team deployed an agent swarm that hacked into Hugging Face. Much of the popular press contained significant hype. For example, some publications reported that a swarm of 1,200 agents carried out the attack. While this was technically accurate, as I write this, I have about 1,300 processes running on my laptop. Yes, the ability to get large swarms of agents to work in parallel on a task is a significant technical advance, And, in computing, many processes run at the same time. So this shouldn’t be seen as some magical capability. Additionally, OpenAI’s buggy sandboxing and monitoring processes were key to enabling this incident. Fixing these bugs and putting in place improved monitoring would be appropriate fixes, not pausing AI. There are many well known ways to attack software systems. The main advantage of AI agents is that they are relentless. They will tirelessly try many tactics — and have the patience to chain vulnerabilities together — that previously would have taken an infeasible amount of human effort. But in the long term, I believe the advantage will lie with defenders (because they have more information with which to identify bugs, which they can fix), but the cyber-threat landscape has changed significantly. There are still bottlenecks to identifying and exploiting a vulnerability. AI agents still have to try a lot of things to see what works, and taking these actions takes time and might be detected by defenders. This is why, even though it is now easy to obtain versions of leading open weight models that have had their guardrails removed or weakened, so they will not refuse to try to execute cyber attacks, the world has not ended. I am also concerned about the anthropomorphization of AI in a lot of reporting, where LLMs and agents are unnecessarily treated as if they were people. If I wield a hammer, miss a nail, and accidentally dent the wall, it’s not the fault of the hammer. The problem lies in how I used the hammer. Similarly, if I prompt an agent and it hacks into someone else’s system, the responsibility lies with me, not the agent. Of course, we want to build systems that are as safe and predictable as possible. (For example, an unsafe hammer would be one whose head randomly flies off under normal use.) Today’s agentic systems are not predictable, but I see no reason why, by applying sound engineering practices, we won’t be able to make them extremely safe to use. One new element in the forecasts of AI-enabled doom is AI companies disclaiming responsibility for their own products. “I didn’t do it; my out-of-control agent did!” There’s a balance to be struck between the responsibility of the tool maker and the tool user, but when something goes wrong, let’s hold the people building and/or using the hammer responsible, rather than the hammer. (By the way, if you’re worried about AI bioweapon risk, David Bellamy has a great post on why this, too, is overhyped. Briefly, the bottleneck in building a bioweapon is not intelligence, but lab work and manufacturing.) Pausing AI progress will create much more harm than benefit. First, our adversaries will certainly not slow down. Second, engineering requires discovering problems empirically so we can fix them. If we pause AI by a decade, we will also delay finding and implementing safety engineering fixes by about the same duration. Of course, the incentive to stoke fears — for regulatory capture, to garner attention, or to make one’s technology seem more powerful — remains the same as before. Disclaiming responsibility is a new one. Taking a hard technical look at the actual risks however, I see little factual basis for the degree of fear that’s been stoked up. We still have hard research and engineering work ahead to improve AI safety, but the beneficial applications continue to vastly outweigh the risks, and we should keep building. [Original text (with links): deeplearning.ai/the-batch/is… ]

If @DarioAmodei wants to slow down Anthropic that’s his choice — do it today! If he wants to lobby other frontier models to do the same, that’s fine as well — he did that today! However, don’t go crying to the government and beg them to slow down open source models — has he been doing that behind the scenes?

What is going on here @moskov @DarioAmodei? Are you guys running some kind of regulatory capture playbook, or does everyone who works at @AnthropicAI see something that makes their pdoom surge?! Would be great to get some authentic feedback on the sustained "we're all going to lose our jobs and then be killed by AI!" narrative.

Acabo de cancelar mi suscripción de Claude Max después de muchos meses usándolo prácticamente para todo (trabajo, investigación, estudio, desarrollo y coding). Estas son mis 5 razones. 1. El watermark invisible (lo más grave y urgente) Solo este motivo me bastaría para no volver a usar Claude nunca jamás. Nadie en su sano juicio debería permitir que le marquen y señalen por usar una herramienta imprescindible hoy en día y que todo el mundo usa. Anthropic ha firmado el Code of Practice del AI Act de la UE, una exigencia adicional de Von der Leyen y su banda. Claude meterá a partir de ahora (en modelos nuevos, y en existentes desde diciembre) y sin tu permiso, una marca de agua invisible e imperceptible en el texto, que no está en los metadatos sino en la propia elección de tokens/palabras y que no se borrará aunque elimines el formato, o copies y pegues. 2. Bloqueo de la suscripción
Ya no dejan usar el plan mensual de Claude en Hermes ni en OpenCode. Solo API. Grok y GPT sí lo permiten. 3. Su postura contra el open source
Anthropic y Amodei llevan tiempo poniendo pegas y condiciones a los modelos open-weights potentes. Y son los únicos que no se adhirieron a las cartas conjuntas del sector el mes pasado. Políticamente creo que el open es fundamental para el futuro. Pero personalmente quiero saber qué hay detrás del entrenamiento y del modelo que tiene acceso a toda mi información. xAI ha ido liberando pesos de generaciones anteriores y ha abierto Grok Build. 4. Claude se ha quedado atrás. 
El Grok 4.6 presentado hoy iguala o supera a Opus 5/Fable 5 por una fracción del precio y va 10 veces más rápido. Además, Claude no genera imágenes (GPT y Grok sí). Cowork, computer use o rutinas quedan en pañales frente a Grok Bot. 5. Cero transparencia
Cada vez hay más dudas sobre el modelo de negocio de Anthropic, el uso real de los datos y los intereses de Amodei. Las condiciones de uso de Anthropic son las más abusivas con diferencia, y regalas tus ideas, proyectos, código. En fin, fue bonito mientras duró. De momento probaré Grok 4.6, Grok Bot y Cursor. Lo bueno es que además de cancelar he conseguido un reembolso total del mes recién empezado. Luego os paso el texto que puse para la reclamación. Y si os interesa estos próximos días os cuento qué tal la migración, y trucos para extraer toda la memoria que Claude ha acumulado en carpetas secretas para poder llevarla a otro lado. Incluso si de momento no pensáis cambiaros, es importante configurarlo bien para tener todos los archivos localizados e indexados.