yawnxyz

they also keep the papers as a pre-edited, free version of the peer reviewed equivalent

> Healthy older adults (n = 123, average age 72.0 years, body mass index 25.8 kg/m2) completed three 45-min supervised exercise sessions per week for 6 months. Participants were randomised to treadmill-based moderate-intensity training (n = 45), or high-intensity interval training (n = 41) or a low-intensity active control condition (n = 37), with individualised heart-rate prescription. Dual-energy x-ray absorptiometry was used to quantify body composition at baseline, and at 3 and 6 months.

ok they didn't even include light/moderate weight lifting as another control, so this is a fairly poorly executed study

basically it comapres hiit with treadmill walking in which case yes, it's slightly more useful, but hiit also causes a lot of damage in a lot of ways

> Author: Deena Mousa

clearly an article sponsored by Big Mouse

> "Bot fatigue sets in when the internet stops feeling honest"

if there's anything worse than LLM-written text, it's websites that rally against LLMs and AI-use, then blatantly just use AI to do the thing they're against

if you're going to be anti-ai, at least don't use it!!

before AI was around we blamed Covid for doing this to us, and now we blame LLMs... and before that we blamed social media. I'm pretty sure this downtrend has been happening for decades.

it would be racist if we started calling it the texas screwworm, but that's kind of redundant

apparently it has to be destroyed and can't be sold

yeah it happens at smaller companies - someone insulted the boss and got sent home (small agency)

layoffs happen, and they're firings "without a real reason" - and no PIPs exist for surviving layoffs

yeah that's true, and it gets murky - because employers will just wait til an employee messes up (which everyone does), and use that as an excuse

> Generically, it's about a unit time of life and how it is spent meaningfully

technology has generally flooded us with more speed, more choice, more entertainment - even the introduction of bicycles caused a similar outrage response, that we're moving too fast and should be slowing down to take in the world around us

the paradox is that choice is both great and awful for us

the one skill to hone / develop in the last couple of decades (way before AI) is the ability to focus, filter, discard, and choose a direction to move in (whether its hobbies, career, apps to build, social media to consume, etc etc etc)

in most states, as long as it's not discriminatory, you CAN fire anyone for any reason

the article is wrong here

self-hosting is fine, but even if you had a $100k god box with opus-level LLMs, you'd still end up grinding it to a halt if you tried running 5-10 parallel inference streams

saying the quiet part out loud

Svelte + Astro is the combo for me - this combo is great for documentation sites and pages w/ selective amounts of interactivity

that's what happens when your only industries are cooking oil, actual oil, and real estate

can you make Superset work on folders that aren't git repositories?

this is why people still enjoy eating at Olive Garden and Chipotle and Sweetgreen

basically the AI-slop version of food, yet still they thrive

the trope is that they actually said the quiet part loudly

Stainless was a fantastic product; every product/service has to start from somewhere

it's nice to have open source implementations so you can extend / borrow / merge / morph features and ideas into something else

my tiny blogs no one reads have been racking up a huge deno deploy and vercel bills ($40-50/mo each) bc I ran them "naked" without a cache or cloudflare or static builds - it didnt matter bc I got like hundreds of visitors a month. they were just hono or whatever api pulling from my backend which could be notion or airtable - super simple, though kind of slow

now I suddenly I have 10k visitors a month hammering my apis and causing massive egress and cpu usage - so i had to get them behind cloudflare and now build everything statically - cut the costs back down from 90+ cpu hours to about 0.2 cpu hours a month

crazy times

(also, all donw w/ claude code's help, or it would have taken a week for me to figure out)

really good observation, and it's interesting what will happen when implementation gets cheaper - designers will get overwhelmed with too many things to review now, probably

though technically your data lives within Bluesky's servers (kind of like your money lives in a bank) so in some ways it's still the bank's money though it's allotted to you

no idea why you were downvoted, but I stopped reading after the first few paragraphs thinking it's almost 100% AI constructed. Very strange article

for our game dev project in undergrad we built an rts game as our capstone, and we used the "slow stuff down" trick - except different systems had different clock/time systems! and so some things were still running while it was paused, which led to some weird side effects (eg money glitch)

i did the thing you're "not supposed to do" and attached everything to a world clock so everything ran at like 60fps in terms of events, so it was a real-time "turn based" system

well that's harness territory! give it the right harness/environment!!

skills can have code bundled with them, including MCP code

I think this speaks to the product release iself

as someone who comes from UX design, and have worked a few years as devrel in AI - people want devrel more than ever. It's now more about exploring the "art of the possible" - to be a guide into uncharted territory.

You're acting as a guide towards product, design, strategy, and other directions that makes sense for the community members (forum, Discord), for your customers and potential customers (as solution architect, building demos and prototypes of where they can develop their product), and generally as kind of eyes and ears of the product community of both AI in general, and in specific niche fields.

For example, I'm now exploring deep into biology, and what can be done to help solve (or alleviate) to make work in biology that was traditionally a slog (paperwork, bioinformatics/pipelining, search, connecting the dots) into something more fun and exploratory.

Devrel isn't going anywhere - it's becoming more product, SA, sales, design, and community. Honestly, this is going to be the decade of devrel.