Author: robert.adlington

if suicide is a self-directed, injurious behaviour, how far can a person be judged to be self-directed when acting under the influence of a purposeful online community? And is a person really choosing freely, Walton asks, when algorithms, which continued to show Aimee content relating to self-harm, power a darkening circle of interest and exposure?…

Avec l’âge, chez l’être humain comme chez la souris, quelques cellules souches se mettent à proliférer de façon massive, écrasant la concurrence en produisant des myriades de cellules identiques. Quelques clones prennent ainsi progressivement le contrôle de la production sanguine – on parle d’« expansion clonale ». Qui plus est, ils produisent de préférence des cellules myéloïdes,…

Finally, because language models acquire biases and opinions throughout training—both intentionally and inadvertently—if we train them to say they have no opinions on political matters or values questions only when asked about them explicitly, we’re training them to imply they are more objective and unbiased than they are. We want people to know that they’re…

Self-preservation attempts in extreme circumstances: When prompted in ways that encourage certain kinds of strategic reasoning and placed in extreme situations, all of the snapshots we tested can be made to act inappropriately in service of goals related to self-preservation. Whereas the model generally prefers advancing its self-preservation via ethical means, when ethical means are…

Hypernormalization captures this juxtaposition of the dysfunctional and mundane. It’s “the visceral sense of waking up in an alternate timeline with a deep, bodily knowing that something isn’t right – but having no clear idea how to fix it”, Harfoush tells me. “It’s reading an article about childhood hunger and genocide, only to scroll down…

A beautiful timeline going back 4.6b years — Read on szupie.github.io/supercontinents/ With source code at https://github.com/szupie/supercontinents

fortune favors the prepared mind. If you’ve developed the habit of paying attention to the things that happen around you and to you, then, yeah, you’ll get hit by lightning. — Read on www.themarginalian.org/2023/09/20/octavia-butler-advice-on-writing/

To build LegoGPT, the Carnegie Mellon team repurposed the technology behind large language models (LLMs), similar to the kind that run ChatGPT, for “next-brick prediction” instead of next-word prediction. — Read on arstechnica.com/ai/2025/05/new-ai-model-generates-buildable-lego-creations-from-text-descriptions/

I was Claude pre-Claude. I once prided myself on how quickly I could write well. Memos, strategy documents, talking points—you name it. I could churn out 2,000 words an hour. That skill is now obsolete. I can still write better than the models, but their speed far outmatches mine. And I know their quality will…

There are whole categories of coding problems that look like this, and LLMs are damn good at nearly all of them. — Read on simonwillison.net/2025/May/13/luke-kanies/