Sign in

Seb

@sebmac.bsky.social
40 followers 59 following 22 posts

Internet is cool but get outside, it gets even better

PostsRepliesMedia
Reposted by Seb
DuckDuckGo @duckduckgo.com · 25/08/2026
1 in 3 AI users have told a chatbot something they kept from trusted people in their lives like a doctor or close friend. We surveyed nearly 2,000 U.S. adults about AI privacy. The results suggest that people are sharing secrets with AI that Big Tech doesn’t keep.
3216
Seb @sebmac.bsky.social · 21/08/2026
Valid critique, it was a good idea to publish this letter
000
Seb @sebmac.bsky.social · 16/08/2026
"Wealthy local coupe" is a massive understatement here. It's been built by one of Poland's richest businessmen. Supposedly built as a gesture of gratitude after overcoming a life threatening illness
111
Seb @sebmac.bsky.social · 08/08/2026
Thanks to the awesome neighbors to the east, Poland has been investigating a lot into cybersecurity for the past few years. It takes time to upgrade everything
000
Seb @sebmac.bsky.social · 02/08/2026
My infant son vomited on our bed when falling asleep. Instead of changing everything I cleaned it up with a towel and dried it with a hairdryer to get him back to sleep ASAP. Just parenthood things
021
Seb @sebmac.bsky.social · 25/07/2026
Interesting study. In any normal country I'd expect light emission after a refinery fire to rebound or increase due to rebuilding effort required around the clock
010
Seb @sebmac.bsky.social · 25/07/2026
That's so sad... Such a great artist, his covers were brilliant, many were poster-worthy
030
Seb @sebmac.bsky.social · 24/07/2026
If PRDs were "written with AI help" I can understand them. Nothing worse in PM these days than a flood of AI slop docs and tickets that stop making sense in the product context after you get past 3rd sentence
000
Seb @sebmac.bsky.social · 22/07/2026
Unfortunately, in many cases very high energy prices stand in the way of that happening sooner.
000
Seb @sebmac.bsky.social · 16/07/2026
I think we'll end up only reviewing fixes to what the LLM did wrong and maybe changes to crucial components. It's risky ofc but no real alternative so far. Long term better models for reviews will help
001
Seb @sebmac.bsky.social · 03/07/2026
That's how a democratic society works - media makes a noise, voters get outraged, politicians notice and adjust the politics to get reelected. It's up to the educated voters to not get baited by propaganda, and sadly we, the voters, are not smart enough to avoid that, as you've shown in your study
010
Reposted by Seb
Gavin Jackson @gavinjackson.bsky.social · 24/06/2026
"Estimates suggest that between a third and a half of their business clients’ use of tokens — a unit of processing power — is for software building" www.bloomberg.com/opinion/arti...
bloomberg.com
Anthropic and OpenAI’s Killer App Might Not Be Enough
As OpenAI and Anthropic PBC prepare to list on public markets and investors sour on tech stocks, an obvious question lingers: Will the business of selling artificial intelligence keep growing? Perhaps...
112
Seb @sebmac.bsky.social · 22/06/2026
Makes sense, at the end of the day the most basic commodity is energy and compute is just energy plus amortization of hardware
000
Seb @sebmac.bsky.social · 17/06/2026
Ah, I see. Thanks for the explanation. Really interesting analysis
200
Seb @sebmac.bsky.social · 16/06/2026
How does that compare to other regions? Since 2000 there were so many technical innovations like LEDs, heat pumps etc. that massively cut household consumption in some areas. Not to mention a lot of our industrial output moved to Asia over that period, this isn't necessarily efficiency gain per se
110
Seb @sebmac.bsky.social · 13/11/2025
How is that supposed to help with context switching lmao. The user of this IDE context switches to their phone? That's not context switching. That's unhealthy lack of focus, which this IDE can only make worse
000
Reposted by Seb
TechCrunch @techcrunch.com · 13/11/2025
Chad: the Brainrot IDE is an actual product that pairs vibe coding with brainrot activities like gambling, Tinder and games.
techcrunch.com
'Chad: the Brainrot IDE’ is a new Y Combinator-backed product so wild, people thought it was fake | TechCrunch
Chad: the Brainrot IDE is an actual product that pairs vibe coding with brainrot activities like gambling, Tinder and games.
3142
Seb @sebmac.bsky.social · 21/08/2025
Google lens on a stick
000
Seb @sebmac.bsky.social · 17/08/2025
In Poland in 2022 misguided changes made it very difficult to figure out the best tax scheme - the demand for accountanting companies surged over 100%. I wonder if similar thing will happen with the tariffs, the rise of new tariff-oriented consultants
000
Seb @sebmac.bsky.social · 12/08/2025
Took this yesterday with my phone camera (Samsung S25, 30x zoom)
A photo of the Moon I took on my S25 camera
010
Seb @sebmac.bsky.social · 12/08/2025
As a Linux user I can confirm this works
linux perpetual energy machine meme
150
Seb @sebmac.bsky.social · 10/08/2025
I think a lot of managers and executives got too excited with the vision LLM companies were selling and this, more than anything, slowed down hiring of junior engineers. It's a pendulum that's pushed out of balance to the extreme by hype and misunderstanding of the topic - it'll move back eventually
020
Seb @sebmac.bsky.social · 08/08/2025
Didn't know I was into wizards until I found this gem of a subreddit
010
Seb @sebmac.bsky.social · 31/07/2025
This really looks like an ironic version of Rollercoaster Tycoon, the art style looks great! Love the idea
080
Seb @sebmac.bsky.social · 20/07/2025
Sycophancy in LLMs is interesting, it might lead you toward a dead end in technical tasks, but seems like it could be dangerous when ppl use ChatGPT like counseling openai.com/index/sycoph...
openai.com
Sycophancy in GPT-4o: What happened and what we’re doing about it
We have rolled back last week’s GPT‑4o update in ChatGPT so people are now using an earlier version with more balanced behavior. The update we removed was overly flattering or agreeable—often describe...
010