I spent a couple of weeks building a model to predict which oncology drugs would advance from Phase 2 to Phase 3. It didn’t work. Walk-forward AUC sat between 0.50 and 0.55 across every feature combination I tried: pathway position, publication sentiment, genetic evidence, novel-MOA flags, sponsor identity. None of them generalised out of sample. But there was a chart I liked too much to let die at the same time. ...
The penguin… Did this go Did it go viral because animals let us talk about ourselves without ego defense? Most viral content today is driven by sparking outrage. In contrast, the penguin felt so calming, it conveys the feeling “I don’t know where I’m going. I just know it’s not there.” Existentialism. Meaning isn’t discovered by following the correct path - it’s created by choosing despite absurdity. For those who built their identity on achieving. For those who reached milestones, and found it left them feeling empty. ...
I have noticed trying to hold myself to the habit of a weekly journal or post gives a different quality to the passing of time. “Modern man sees himself as irreversibly situated in historical time, while archaic man defended himself against time by periodically abolishing it.” — Mircea Eliade, The Myth of the Eternal Return Things found this week… 1. Czechoslovak Embassy in Stockholm, Sweden 2. Aristotle on how tyrants maintain power ...
A collection of others cool and unique imprints on the world/internet… Cool things found this week: Barcelona, 1906. The Sagrada Família basilica in the background, 24 years into its construction. Midwinter evening by a forest road, Dalarne (1913) Peder Mørk Mønsted (1859–1941) L’Empreinte de Dieu dans le monde quantique by Yves Dupont Finding out Frank Gehry designed a yacht
Trying to write a less cringey post that feels more myself… I think my posting is about to go through the awkward transition most of us should have had as teenagers: putting thoughts out into the world before they’re fully formed, unsure if anyone is reading, finding our ‘voice’, making the necessary grammatical and conceptual mistakes… ideally so my future self doesn’t have to. As a teenager I was far too concerned with what strangers might think to run those experiments in public. So it seems I’m doing it now in my 20s, a little later than planned. Somewhere along the way that curiosity got postponed, but lately I’ve become mildly obsessed with the idea of leaving some kind of trace on the internet. Not a brand, not something polished, just something honest. Just evidence of thinking in progress. If anything coherent emerges from the rambling, that will be a bonus. For now, I’m starting here. And, I’m starting safe, with a collection of others cool and unique imprints on the internet… ...
This is pt.2. Pt.1 covers the raw materials overview. The first post catalogued what AI hardware is made from. This one tries to answer a more specific question: where in the supply chain does the actual geopolitical leverage sit? The assumption most people carry is that it sits at extraction. For almost every mineral that matters for AI hardware, China’s share of refining and processing is substantially higher than its share of mining. Often dramatically so. ...
I wanted to think about potential metrics for modelling POS - the following is a living document, open iterative work on this subject. Related posts will update on findings, failures and further ideas to be tested! Predicting probability of success (POS) in biotech should start with biology. Evaluating the mechanism of action (MOA) of therapeutic targets, differentiating between upstream and downstream interventions, and, how the position of a target within a biological pathway influences POS. Most models overweight market size, management quality, or phase transition statistics and underweight whether the mechanism actually has durable control over the disease system. The key question is where the target sits within the network. ...
“ The demonstration that no possible combination of known substances, known forms of machinery and known forms of force, can be united in a practical machine by which man shall fly long distances through the air, seems to the writer as complete as it is possible for the demonstration of any physical fact to be.” - Simon Newcomb (1906) J. Storrs Hall’s book ‘Where Is My Flying Car?’ asks why certain anticipated technologies haven’t become mainstream despite our technological capabilities. Hall uses the flying car as a metaphor to discuss innovation stagnation that has plagued since the 1970s, a period he refers to as the “Great Stagnation.” ...
AI is not purely digital infrastructure. It is an industrial system. Last year, went down the rabbit hole on earth materials and how they fit into the theme of supply chain demand, particularly with the growth of compute required AI. Defense contracts heavily depend on lithium, especially for electronic components and warfare applications. Currently, the US mainly rely on outsourcing lithium from countries such as China and Australia. The interesting realisation is that the bottlenecks increasingly look less like traditional software constraints and more like industrial coordination problems. ...
I want to talk about a rabbit hole I have fallen down since reading a paper on The Promises and Predicaments of Federated Learning in Healthcare. Last year, I had the privilege of working with an awesome team, we explored ideas applying machine learning to tackle data interoperability challenges within our healthcare system. The problem: disparate data formats, strict privacy regulations, and the sheer volume of sensitive patient information scattered across multiple institutions. Our thesis was that the near-term impact of data interoperability is obvious, time saved and admin (huge burden to the NHS) reduced. We validated this with our study across thousands of hospital reported EHR systems, and with firsthand NHS clinical experience in the team. But the bigger picture? The data unlock. ...