2 days until the Wrocław Night Half Marathon. I picked up my race kit and enjoyed a lovely Swedish fika with cardamom buns. I have generated enough data for both Garmin and Trait so now they more or less agree on my upcoming performance: Trait predicts a half marathon in around 1:41:55, while Garmin estimates 1:41:05. Since both assume a distance of 21.1 km and I’ll likely cover 21.25 km, I’m aiming for 1:42:35.

Only 4 days left… time to take it easy :)

Today I got rid of my Instagram account… in two weeks or so I will delete my Facebook account!!

No doubt Garmin is a perfect execution platform but planning and predicting is not its strong suit. My running app says I can run half marathon in about 1:42. So far my best (from month ago) was 1:44:20 and Garmin prediction is 1:48.

Here’s the thing! My last half marathon was 1:44:20 though it was tough race with two long uphill sections. Then I laid my hands on Trait, a new training app (it popped up when I signed up for Dresden HM). Super simple: you set your goal race, date, and effort level, and it does all the planning for you. So my next race is Wrocław Night HM in 18 days. Based on my last time, I was aiming for 1:42:50–1:43:10... but Trait predicts I can hit 1:40:40–1:41:40 on race day. I find it hard to believe, but okay - let’s see! Now here is the trouble - zoning. Wrocław expects 32,000 runners, and the start zones are… weird: Green zone: 1:30–1:41 Next zone: 1:41–1:50 (that one is probably hold 70% of runners) So… which one do I pick? If I choose green then it is a bit of pressure to make sure I keep 4:59 pace. If I go for the next I might drown in a crowd of people and put up extra meters to go round them… for the moment green looks more reasonable.

Is AI the nail in the coffin of capitalism? Saturated markets cannot absorb more goods, and companies cannot grow indefinitely in a world of constrained resources and population. At the same time, with the rise of automation, a handful of skilled individuals with computers can achieve what once required large organizations. If people don’t work, who will buy? “If AI displaces human workers faster than the economy can reabsorb them, it risks eroding the very consumer demand firms depend on. We show that knowing this is not enough for firms to stop it. In a competitive task-based model, demand externalities trap rational firms in an automation arms race, displacing workers well beyond what is collectively optimal.”

arXiv.org

The AI Layoff Trap

If AI displaces human workers faster than the economy can reabsorb them, it risks eroding the very consumer demand firms depend on. We show that kn...