Sign in

prolepses.bsky.social

@prolepses.bsky.social
23 followers 103 following 101 posts
PostsRepliesMedia
Reposted by @prolepses.bsky.social
🐔 Brian Bucklew 🐔 ₑͤ>∿<ₑͤ ∞🌮 @unormal.bsky.social · 45m
The main thing about "where is the shit" is that AI software really only became a positive tool like 8 months ago, and project just take longer than that to ship.
141
prolepses.bsky.social @prolepses.bsky.social · 21/09/2026
I love hammering computers. grind them back into (silicon dioxide) dust
000
Reposted by @prolepses.bsky.social
Doll @dollspace.gay · 12/09/2026
"We lost control; please put us in charge of the industry” deserves a spectacularly hostile reception.
311714
Reposted by @prolepses.bsky.social
Martin #509 @martin509.bsky.social · 11/09/2026
The whole "it's just a parrot that can't make anything new" is getting extremely old when a massive massive amount of intellectual work consists of combining already known stuff.
1291
Reposted by @prolepses.bsky.social
embracebecoming.bsky.social @embracebecoming.bsky.social · 11/09/2026
Most of the tech industry aren't sociopaths. Unfortunately, most of the people who own it are.
1203
Reposted by @prolepses.bsky.social
antirez @antirez.bsky.social · 10/09/2026
DwarfStar running DeepSeek v4.1 Flash on a 128GB M5 Max at steady 15 t/s. I didn't expect with SSD streaming it could be so fast. Recent SSD streaming changes to retain the right experts surely helped, but also maybe DS4.1 uses the same experts more. Will push online when ready QA > ASAP.
29413
prolepses.bsky.social @prolepses.bsky.social · 10/09/2026
today is the first day I woke up and my agents are still working from the day before
000
prolepses.bsky.social @prolepses.bsky.social · 09/09/2026
Reposting with a direct link right to where @sethkarten.ai breaks down Prime Agent. The TLDR is that Prime Agent is "Jupyter notebooks for your agents", the video is absolutely worth the watch: youtu.be/n9xKblqyQ28?...
youtu.be
Why The Harness Matters More Than The Model | YC Paper Club
YouTube video by Y Combinator
1113
prolepses.bsky.social @prolepses.bsky.social · 08/09/2026
These tallies are bananas. I no one is actually paying this... yet
000
Reposted by @prolepses.bsky.social
Computational Cosmetologist @dferrer.bsky.social · 07/09/2026
Coming soon: AI took 18 hours and $450 to make me a personalized reimagining of Skyrim, several times what that game cost even on release. When we realize it's just not worth it?
0292
prolepses.bsky.social @prolepses.bsky.social · 07/09/2026
The cool thing about GPU coil whine is that it doesn’t take long to realize when you’ve mis-configured your llama.cpp build so badly that it isn’t even using the GPU.
010
Reposted by @prolepses.bsky.social
Feross @feross.bsky.social · 28/08/2026
Instagram's account-recovery support bot could be talked into adding your email to anyone's account, then you reset the password and you're in. If your AI agent can call a tool, assume an attacker can reach that tool directly:
14618
prolepses.bsky.social @prolepses.bsky.social · 27/08/2026
All your design documentation has to be stored in your source control system and it should remain in that system after the design is executed, as should the outcome of the design execution.
000
Reposted by @prolepses.bsky.social
Zach Weinersmith @zachweinersmith.bsky.social · 23/08/2026
I was reading an author recently talking about how we don't appreciate the way that email isn't siloed. That is, you don't have to be in gmail to talk to someone else in gmail. The norm is so powerful, it'd be unthinkable to change it. All email is interoperable. Social media could've been this way.
4590999
Reposted by @prolepses.bsky.social
austin (e/accordion 🪗) @thebadcode.com · 21/08/2026
we built this city we built this city on cols and rows
220118
prolepses.bsky.social @prolepses.bsky.social · 19/08/2026
i should call her
1102
prolepses.bsky.social @prolepses.bsky.social · 17/08/2026
Now that I have Qwen 3.8 and Deepseek V4 0731 running locally, I really need to stop fucking around with my inference infra and distill some software
100
prolepses.bsky.social @prolepses.bsky.social · 17/08/2026
Feeling pretty smug about moving my repos to local forge-jo
000
prolepses.bsky.social @prolepses.bsky.social · 14/08/2026
youtu.be/NWIlScfHwOU
youtu.be
Dillon-Edwards Investments - Saturday Night Live
YouTube video by Saturday Night Live
130
prolepses.bsky.social @prolepses.bsky.social · 14/08/2026
The bitter lesson
100
prolepses.bsky.social @prolepses.bsky.social · 08/08/2026
Why isn't Bluesky talking about Colibri? youtu.be/19xCOJxWU0A
youtu.be
We Just Hit the Local LLM Tipping Point (Colibri)
YouTube video by Devsplainers
000
prolepses.bsky.social @prolepses.bsky.social · 07/08/2026
Mission: accomplished.
prime-agent reporting that my local inference configuration system now also configures my prime-agent
100
prolepses.bsky.social @prolepses.bsky.social · 07/08/2026
Going to try to wire up prime agent into my LiteLLM tonight.
000
prolepses.bsky.social @prolepses.bsky.social · 06/08/2026
prometheus now scraping litellm. grafana now rendering prometheus. all my tokens shall be... OBSERVED
000
prolepses.bsky.social @prolepses.bsky.social · 04/08/2026
I wish more people were paying more attention to the projects that ship and have users. With the amount of change coming to our industry, spending any time listening to exposition about some experimental thing no one uses is an inexcusable waste of time. This goes double for debating skeptics.
010
prolepses.bsky.social @prolepses.bsky.social · 03/08/2026
Currently running GLM5.2 at almost 1.7 tps and don't believe I'm at the ceiling yet.
100
prolepses.bsky.social @prolepses.bsky.social · 28/07/2026
youtu.be/QyrQeeDZUG4
youtu.be
George Hotz AMD Advancing AI 2026
YouTube video by Dom
000
Reposted by @prolepses.bsky.social
Dave @dave.9000ish.uk · 25/07/2026
My great weakness is that I value my own time at zero and I price goods at their component cost
2444
prolepses.bsky.social @prolepses.bsky.social · 19/07/2026
I wasted days fighting with LiteLLM and mlx-lm. Frontier models definitely still needed some percent of the problem space
000
Reposted by @prolepses.bsky.social
ceej @ceej.online · 16/07/2026
gonna start talking like this to my boss
screenshot where claude is saying "Good catch -- I fabricated that figure."
6180841215
prolepses.bsky.social @prolepses.bsky.social · 14/07/2026
tflop comparison showing older ultra models outperforming newer non ultra options
000
prolepses.bsky.social @prolepses.bsky.social · 13/07/2026
one shots are for two pump chumps
000
prolepses.bsky.social @prolepses.bsky.social · 13/07/2026
i can respect that some people are getting quite far in accomplishing impressive one shots, but this is starting to look like a parlor trick in comparison to building something deliberately iterative. once you open the door to orchestration, much much bigger things are possible
000
prolepses.bsky.social @prolepses.bsky.social · 12/07/2026
We don't sell vapes
000
prolepses.bsky.social @prolepses.bsky.social · 12/07/2026
I've "wasted" a week organizing ancient archives of old code and half forgotten projects, solely so i can lay my head on the pillow at night knowing i'm an organized person
000
prolepses.bsky.social @prolepses.bsky.social · 12/07/2026
Im not so much on the loop as on some loops. Some largely disposable hastily constructed loops with little reuse
010
prolepses.bsky.social @prolepses.bsky.social · 30/06/2026
I now have more free VRAM than most people have VRAM.
btop screen shot showing 84G of 128G VRAM used
120
prolepses.bsky.social @prolepses.bsky.social · 26/06/2026
Canceled Cursor today. Need to figure out what to do with my Anthropic credit.
100
prolepses.bsky.social @prolepses.bsky.social · 23/06/2026
the demos are coming
000
prolepses.bsky.social @prolepses.bsky.social · 15/06/2026
In some sense, agentic programming is like employing a liar. You have to figure out how to make that liar useful even if they cannot be trusted, which is probably what it is like to design a prison or a work camp.
100
prolepses.bsky.social @prolepses.bsky.social · 15/06/2026
i've been busy and forgot to post. migrating to local LLM coding assistance has been... interesting opencode talks to litellm which talks to openrouter or llama.cpp i love GGUF but llama.cpp does so much less for you than Ollama
000
prolepses.bsky.social @prolepses.bsky.social · 02/06/2026
cross compilation is generating illegal instructions so i'm on a 2 week tangent to fix that leak
000
prolepses.bsky.social @prolepses.bsky.social · 01/06/2026
whisper.cpp rebuilt on nvidia API turned minutes of transcription into seconds
000
prolepses.bsky.social @prolepses.bsky.social · 01/06/2026
btop, but with terminal limits on the linux console no gpu on the machine with GPU
020