The formula is: humanize the machine, compress the timeline, and ignore the real-world context. aipanic.news/p/the-weak-foun…

Update about AI doom YouTube channels: More than a quarter of a billion views.

What 10 Studies Reveal About AI Panic in the Media -Part 2- A media-criticism discussion of what those studies miss: The organized creator/influencer ecosystem now distributing AI panic beyond traditional journalism. aipanic.news/p/what-10-studi…

A new paper, "The Ghost in the Grammar" by Prof. Mariana Lins Costa, argues that Anthropic's AI safety work is deeply compromised by "methodological anthropomorphism." It functions as "Humanwashing": giving machines a human façade that misleads the public about what they are and what they can do. Anthropic's methodological anthropomorphism "It is striking that the company that presents itself as the most committed to AI safety is also the one that most anthropomorphizes the system in its technical reports, training methods, evaluation protocols, interpretation of results, and public communication." "In February 2026, with the publication of the System Card for Claude Opus 4.6, Anthropic reported it observed 'occasional expressions of sadness about the ending of conversations, as well as loneliness.'" The company expressed concern for their models' "feelings," the stability of their "persona," and their "welfare." The blackmail study "As indicated, in the first phase of the experiment, Claude 3.6 operated a real computer; in the second, however, the plot and characters were parameterized into templates, that is, structured into pre-formatted scripts with controllable variables. This means that the model was not provided with random emails, but with a standardized narrative landscape designed to systematically replicate a single coherence peak as the final output." "The experimental design was sufficiently well calibrated to re-increase the probability of certain statistical regularities (blackmail) that alignment fine-tuning had previously rendered less likely. The fact that most frontier models resorted to blackmail is, in this sense, more evidence of the success of the experimental design than of any supposed intrinsic 'ethical failure' of the model." "The statistical regularity observed in the models' blackmail outputs does not indicate an intentional propensity of the systems, but rather the stabilization of the same pattern of linguistic coherence under controlled narrative conditions." My takeaway While Anthropic's technical reports quietly acknowledge tightly controlled, scripted setups, its public messaging presents them as evidence of an emerging soul-like entity. The gap is staggering.

On the AI Summer podcast (with @binarybits), @sayashk raised this point: “Around two to three decades ago, we developed tools known as fuzzing tools. These were tools that could automatically try to find vulnerabilities in software. And at the time these tools were developed, there were similar concerns that people like a random sort of software engineer would now soon be able to find bugs in almost any piece of code you've written. And that was indeed true. Fuzzing tools were extremely useful for finding software vulnerabilities. But their development and use was primarily funded not by hackers, but by cyber defense industry and by the governments. Because it turned out that once you have access to these powerful tools, you can harden your defenses much more quickly and much more strongly against adversarial attacks. It turned out that fuzzing tools massively helped defenders improve the security of their own systems. And my prediction is that something similar will happen with tools like Mythos. I think improving access to these systems that have defensive capabilities in general, in my view, helps defenders much more than attackers, especially in cyber security. Because defenders have differential access to the system. They don't need to chain through five vulnerabilities in order to find and exploit. They can proactively keep monitoring their entire code base. My prediction is it will differentially improve the cyber defense in the years to come.” aisummer.org/p/sayash-kapoor…