PhD student under Pierre-Luc Bacon and Doina Precup. I split my time between research and building.
Research
Working on questions like:
- Do we really need silicon-based computation?
- Can we build interfaces that adapt to users rather than the other way around?
- Can we train bigger neural networks in a distributed fashion?
Tools like experiment_buddy aim to make AI researchers’ workflows smoother.
Book a chat: calendly.com/manuel-delverme/15min
Quantified Self
I track everything. 26 data sources spanning 13+ years: Spotify plays, GitHub commits, browser history, location data, messages, calendar events, finance transactions, health metrics.
The goal: identify causal factors in personal productivity through longitudinal self-experimentation. Most productivity advice relies on anecdote, survivorship bias, or small-N studies with poor ecological validity. With 13+ years of continuous measurement, I can run proper before/after analyses and control for confounds.
What I’ve Found
Some findings confirm conventional wisdom:
- Momentum matters (yesterday’s output predicts today’s)
- Morning focus correlates with higher output
Some findings contradict it:
- More meetings correlate with more commits, not fewer
- Late nights correlate with peak productivity (when you’re in flow)
Some findings surprised me:
- Spotify skip rate >20% predicts a 55% drop in next-day commits
- Hands-up/eurodance music correlates with 6+ commits/day
- Media volume, not timing, predicts next-day crashes
The Framework
I think about life optimization as maximizing:
V_total = ∫₀ᵀ V(t) dt
Where V(t) = daily value, T = lifespan. Health extends T. Everything else optimizes V(t).