







A `sleep` with progress
Nicolas Mattia – Keep your Mac awake with caffeinate
This explains how to keep a Mac from going to sleep using a simple command

Matt Pocock on Twitter / X
There's an AI coding approach that lets you run seriously long-running AI agents (hours, days) that ship code while you sleep. I've tried it, and I'm not going back.It's the Ralph Wiggum approach. Here it is in a nutshell:Run a coding agent with a clean slate, again and again…— Matt Pocock (@mattpocockuk) January 4, 2026
Sleep: Educational Impact and Habit Formation
In a field experiment among US university students, incentives for sleep increase both sleep and academic performance. Our primary intervention pairs bedtime reminders with morning feedback and immediate rewards for sleeping at least 7 hours on weeknights. The intervention raises the share of weeknights with 7+ hours of sleep by 26% and average sleep by 19 minutes during a 4-week treatment period, with posttreatment effects of 8 minutes. Immediate incentives outperform both delayed incentives and reminders/feedback alone during treatment but have no distinguishable impacts posttreatment. The primary intervention improves average semester course performance by 0.075–0.089 grade points, a 0.10–0.11 standard deviation increase. Our results demonstrate that incentives targeting sleep can be a cost-effective approach to improve educational outcomes.
The Mythical Agent-Month – Wes McKinney
Like a lot of people, I’ve found that AI is terrible for my sleep schedule. In the past I’d wake up briefly at 4 or 4:30 in the morning to have a sip of water or use the bathroom; now I have trouble going back to sleep. I could be doing things. Before I would get a solid 7-8 hours a night; now I’m lucky when I get 6. I’ve largely stopped fighting it: now when I’m rolling around restlessly in bed at 5:07am with ideas to feed my AI coding agents, I just get up and start my day.

Language Models Need Sleep
View recent discussion. Abstract: Transformer-based large language models are increasingly used for long-horizon tasks; however, their attention mechanism scales poorly with context length. To handle this, we study a sleep-like consolidation mechanism in which a model periodically converts recent context into persistent fast weights before clearing its key-value cache. During sleep, the model performs $N$ offline recurrent passes over the accumulated context and updates the fast weights in its state-space model (SSM) blocks through a learned local rule. During inference, this shifts extra computation to sleep while preserving the latency of wake-time prediction. We test our method on controlled synthetic tasks, including cellular automata and multi-hop graph retrieval, as well as a realistic math reasoning task, on which a regular transformer as well as SSM-attention hybrid models fail. We then show that increasing sleep duration $N$ for our models improves performance, with the largest gains on examples that require deeper reasoning.
Replacing Cron Jobs With systemd Timers | Mjanja Tech
systemd has a cron replacement, and it’s better.
Life Cycle - Track Your Time App - App Store
Download Life Cycle - Track Your Time by Sleep Cycle AB on the App Store. See screenshots, ratings and reviews, user tips, and more apps like Life Cycle - Track…
autosuspend 11.3.0 documentation
autosuspend is a daemon that periodically suspends a system on inactivity and wakes it up again automatically in case it is needed. For this purpose, autosuspend periodically iterates a number of user-configurable activity checks, which indicate whether an activity on the host is currently present that should prevent the host from suspending. In case one of the checks indicates such activity, no action is taken and periodic checking continues. Otherwise, in case no activity can be detected, this state needs to be present for a specified amount of time before the host is suspended by autosuspend. In addition to the activity checks, wake up checks are used to determine planned future activities of the system (for instance, a TV recording or a periodic backup). In case such activities are known before suspending, autosuspend triggers a command to wake up the system automatically before the soonest activity.
I couldn’t sleep, then I tried cognitive shuffling | Psyche Notes to Self
The usual advice didn’t help my sleep problems, but then I tried an exercise that kickstarts the sleep-onset control system

What Did Ancient Humans Do at Night?
How Cycling Solved Sleep | Defector
AURILLAC, France — Is the Tour de France the most competitive sleep environment in sports? “Probably, yeah,” answered Dr. Jon Greenwell, EF-Education EasyPost’s Head Doctor when I asked him. Every athlete needs to recover, though the demands on professional cyclists at the biggest race in the sport are unique. Riders have to be fresh every…

Utkarsh on Twitter / X
Another reason why you want an ambient memory agent is TIME. Your other coding agents understand time differently depending on who or which company writes them. Your machine however should live and breathe with you in your timeline. With Claude, I often find myself begging… pic.twitter.com/XVAvzciplY— Utkarsh (@saxenauts) May 4, 2026


Timeback - AI-Powered Education | 2 Hours Daily Academic Mastery
Revolutionary AI-powered EducationOS helping children master academics in just 2 hours per day. Personalized learning with 10x faster gains. Join the waitlist for 2026 launch.

Beautifully timed with @sri.xyz's excellent updates to @goals.garden, I've shipped a little project to add two-sync between it and your Every Day Calendar! Get you a QT Py ESP32-S3 and #ATProto all the things! Blog! byjp.me/posts/everyday-calendar/ Code! tangled.org/byjp.me/every-day-calendar/
Goals Garden
🌱 Minor updates : - Use your favorite favicons for your calendar - No more EXIF data for your goal photos - ALT Text option for goal photos - Published lexicons lexicon.store?q=garden.goals - Goal completions to include at-uri fields (along w migration utility to migrate your old completions)