Substrate

A digital garden of notes, ideas, and reflections from Doug Belshaw.

Back to the surface

June 2026

110 notes
notes updated 2 min read Untended

European urban birds flee approaching women sooner than approaching men

AI summary

European urban birds flee about 1 meter sooner when approached by women than by men, a pattern consistent across multiple cities. The cause of this difference is unknown, and it contradicts the expectation that birds would perceive men as more threatening due to historical hunting roles.

…
notes updated 2 min read Untended

On the power of 'nutraceuticals'

AI summary

A meta-review found that several nutraceuticals (omega-3s, SAMe, curcumin, saffron) either enhanced the effects of antidepressants or performed better than antidepressants alone for adult depression, with side-effect rates similar to placebo. The review highlighted that adjunctive use of these supplements often provided superior efficacy compared to either treatment alone.

…
ideas updated 1 min read Untended

Comments enabled!

I've enabled comments here on Substrate. Given that I haven't told many people this space exists – and the dearth of comments I've received in the last decade – I'm not expecting much. But that's not to say I don't want and encourage them!

…
notes updated 1 min read Untended

Responsible by Design

AI summary

The Open University’s Responsible by Design framework (v2) addresses ethical challenges of generative AI in education through six pillars: Bias, Sustainability, Exploitation, Access and Equity, Transparency, and Opting Out. It provides reflective prompts and actionable solutions to help educators make informed, responsible AI choices. The note expresses skepticism about whether the framework will have significant practical impact.

…
reflections updated 1 min read Untended

On 'stability'

AI summary

Dave Gray argues that having a job only feels stable but is in fact one of the least stable arrangements. The note agrees, pointing out that the risk of self-employment is visible, while the risk of a job is hidden.

notes updated 2 min read Untended

Tacit knowlede and domain expertise in a world of LLMs

AI summary

LLMs require tacit knowledge and domain expertise to produce useful results, as domain experts can verify correctness even if they can't code, while generalist engineers can build but lack the "ground truth" to judge domain-specific outputs. The binding constraint has shifted from building to verifying what is correct.

…
notes updated 1 min read Untended

Terminus theme

AI summary

Terminus is a dark retro theme for the Zola static site generator, hosted on GitHub by ebkalderon. The note points to the repository and its inclusion in an Are.na collection focused on a field station aesthetic.

notes updated 2 min read Untended

On 'dopamine fracking'

AI summary

The note explores the concept of "dopamine fracking" as a metaphor for online culture that prioritizes addictive dopamine hits over complexity, nuance, and beauty. The author describes how this leads to homogenized, commodified culture and shares their personal practice of reducing such stimuli by deleting channels, uninstalling apps, and setting boundaries.

…
notes updated 2 min read Untended

AI and corporate profits

AI summary

The note argues that AI agents may reduce corporate profits by eliminating inefficiencies that companies currently exploit, such as customer inertia and lack of optimization. However, it counters that human irrationality and stock market behavior make this outcome uncertain, as seen in past recoveries from tech stock crashes.

…
notes updated 1 min read Untended

Bo Burnham is a genius

AI summary

The note describes Bo Burnham's work as a genius mix of comedy, storytelling, revelation, and music, comparing a new video favorably to his pandemic-era special Inside. The author was impressed by Burnham's creativity in both pieces.

…
reflections updated 1 min read Untended

Just get it written

AI summary

The author reflects on using mindmaps for past projects and estimates their zine will be about 6,800 words. They decide to follow Warren Ellis's advice to create a deliberately bad first draft to stop overthinking.

notes updated 1 min read Untended

On "mattering" in a philosophical sense

AI summary

Rebecca Newberger Goldstein defines mattering as being deserving of attention, particularly self-attention, which makes humans normative, value-seeking creatures. She proposes four empirical categories of "mattering projects": socializers, transcenders, heroic strivers, and competitive matterers. Goldstein closes with the normative claim that the very desire to matter demonstrates that we already do, so any project based on denying others' worth is mistaken.

…
notes updated 2 min read Untended

The emotional aspect of burnout

Ky Decker has written a really interesting post about quitting his tech job. Yes, the proximal cause is AI, but once you get past that, there some interesting points on productive friction, values, and burnout.

No matter how rapidly technology changes, I am coalescing around some core beliefs:

…
notes updated 1 min read Untended

How to make a living

Jason Zweig's father was a smart guy:

There are three ways to make a living:

  1. Lie to people who want to be lied to, and you’ll get rich.

  2. Tell the truth to those who want the truth, and you’ll make a living.

  3. Tell the truth to those who want to be lied to, and you’ll go broke.

    …
notes updated 1 min read Untended

A recipe for trouble

AI summary

The note reports that 13.5% of young people are NEET (Not in Education, Employment, or Training), with a sharp decline in entry-level jobs and hospitality vacancies halving in four years. Since the pandemic, NEET rates for men aged 16-24 have risen 40% versus 7% for women, contributing to feelings of worthlessness and attraction to far-right online influences.

…
notes updated 3 min read Untended

The default option

AI summary

Default bias describes the tendency to stick with pre-set options due to friction, loss aversion, implied endorsement, and cognitive cost. This effect dramatically shapes outcomes in areas like organ donation, retirement savings, and privacy settings. The article distinguishes between endogenous defaults (system-set) and exogenous defaults (social or situational norms).

…
notes updated 3 min read Untended

AI ventriloquism

AI summary

A manager used an AI assistant to mediate workplace communication, forcing the author to front-run his own comments to avoid being misrepresented by the AI. The author argues that LLMs mimic thinking without authentic expertise, and their inability to distinguish worthwhile difficulty from unnecessary friction risks devaluing deep thought and human dignity.

…
Other notes link here
notes updated 1 min read Untended

Embodied cognition

AI summary

This note presents Rachel Botsman’s argument for embodied, firsthand experience over abstract machine logic, drawing on a personal visit to Charles Darwin’s house. It contrasts reflective learning through curiosity and play with productivity-driven or performative motives, and warns against relying solely on the internet or LLMs for understanding.

…
Other notes link here