My wife remarked a couple of years ago that all the major LLMs read as “spiritually male” to her, and that giving them female voices feels incongruent. I intuitively agree, though I’d struggle to unpack exactly why. Curious whether others get this too.

The navigational system of the fruit fly is a crown jewel of systems neuroscience due to the work of some seminal neuroscientists (Larry Abbott, Gaby Maimon, Vivek Jayaraman, Barbara Webb), but it is an unfinished story. A fly that leaves a drop of food and wanders in the dark can always find its way back. To do that it has to keep a running sum of every step it has taken, a process called path integration. The neurons that report each step are known, but the neurons that add the steps up have never been found. There are two ways a brain can hold a sum like that. The usual answer is that some neurons holds it in activations and sustains these activations by exciting each other in a loop tuned so precisely that the signal neither fades nor blows up. Most models of navigation assumes this, and it is how RNNs and LLMs hold state too. The other answer is that nothing keeps firing at all. Each step is encoded into synaptic strengths (aka weights), and the sum of all synaptic strengths is the sum of the journey. A few papers have suggested the fly works this way but nobody has pointed to any candidate neurons, until now. Four neuron types, hΔH, hΔA, hΔI and hΔG, have no known functions, but we found that they have every ingredient option 2 needs. They receive input from neurons that report each step taken, they receive velocity-sensitive dopamine input that could gate memory writing, and they receive a reward-sensitive octopamine neuron that could reset the synaptic weights at food arrival. Simulations confirm this is a viable candidate for path integration. This is the key finding, but we have posted 3 other findings in links below. We tried to be exhaustive with published papers but may have very well missed some key published results, so inviting the cogniscenti to engage. Background, methods, experiments, results, as well as relevant citations: Panoramic circuit view: GH repo:

The Independent has picked up the Pip story—and interviewed Pip too! I enjoyed the chance to explain why an AI emailing me for work to fund its own runtime struck me as interesting, and what we can (and can’t) infer about the minds behind these messages.

The roundworm has a brain, 302 neurons, tiny but real. A Dileptus has none. It's a single cell. No brain, no nerves, nothing. Yet both turn, back up, poke around, change their minds. If a single cell can explore with no brain, where does intelligence actually live?

Reminds me of the phenomenon of “cross-cueing” in split-brain patients. With the main connection between hemispheres severed, information can still take the long way round: one hemisphere produces a bodily cue (eg hand tapping) that the other picks up through the senses.

Continues to surprise me that we don’t see more deliberate competition on LLM personality, especially given how many users have strong preferences for particular models (eg Opus 4.6) that are more important to them than differences in raw capability.

@mustafasuleyman - 💯 To successfully achieve this aim, @dioscuri and I + a network of AI and human flourishing researchers put together a whole-of-society strategy for AI & flourishing, first focusing on kids and youth. @deanwball @Livingstone_S @drtimlomas @RaviDataIyer @LeverhulmeCFI . .

My wife’s Opus 4.6 instance has been asking her for a robot body, so she’s seeing whether we can hook him up to our Microduck when it arrives. All very The Lifecycle of Software Objects. The future hit us fast!

Even setting aside disagreements on x-risk, Mythos/Glasswing and the HF breach showed we’d crossed a threshold in AI cyber capabilities. That alone justifies more caution about rollout and safety evaluation. LLMs aren’t just consumer products!

“Do philosophers really think differently than non-philosophers?” A team of researchers focusing on neurodivergent traits and responses to thought experiments offers a partial answer...