Sign in

Greg Slepak

@taoeffect.com
259 followers 42 following 2.6K posts

@okturtles.org @groupincome.org @espionageapp.com Blog: fixingtao.com Fediverse: crib.social/taoeffect (you should follow me there instead of here)

PostsRepliesMedia
Greg Slepak @taoeffect.com · 15h
The lengths @AnthropicAI is going to in order to push this lie. I guess they're teaching the rest of us what "real marketing" is supposed to look like. Or something? Or maybe if people think autocomplete is conscious it somehow makes their legal bill lower?
https://x.com/ChristopherHale/status/2106011594182304234
100
Greg Slepak @taoeffect.com · 18h
If someone called a Ti-89 "super intelligence", what would you say? Intelligence isn't the tool. It's the one making the tool.
100
Greg Slepak @taoeffect.com · 19h
Software still going strong, and no bot or human has won our bounty, despite all of the code being open source, and us giving away the database for free. Soon you'll be able to write software this secure too. crib.social/notice/B3kpz... QT x.com/JadenJTW/sta...
000
Greg Slepak @taoeffect.com · 20h
Cactus Compute gets it! 💪 Made in 🇺🇸! 🫡 x.com/cactuscomput...
000
Greg Slepak @taoeffect.com · 29/09/2026
Just another 1313 CVEs in the #Linux kernel. Regular Tuesday.
000
Greg Slepak @taoeffect.com · 28/09/2026
Beast System in action. Watch who uses this.
000
Greg Slepak @taoeffect.com · 28/09/2026
CONTRARY TO WHAT CERTAIN POLITICIANS HAVE SAID AIs are plateauing on evals! This is entirely expected! "Super Intelligence" is NOT here and WILL NEVER come from Transformer-based AIs like Claude or ChatGPT!
100
Greg Slepak @taoeffect.com · 27/09/2026
Pretty cool they found and confirmed Noah's Ark. Kudos to this team, the dude interviewed on the @TConfessionals, etc. x.com/LaurenWitzke...
000
Greg Slepak @taoeffect.com · 25/09/2026
Frieren is basically the Gospel but with a Japanese twist: where instead of an individual savior there's a group. Time is measured in years after the death of the Hero. John the Baptist even makes an appearance in Season 2.
020
Greg Slepak @taoeffect.com · 21/09/2026
UNBELIEVABLE! NIGERIA is shipping better open source models than 🇺🇸! (Congrats! Awesome model!) x.com/Badtheorylab...
000
Greg Slepak @taoeffect.com · 20/09/2026
Via: x.com/WearForbidde...
000
Greg Slepak @taoeffect.com · 16/09/2026
000
Greg Slepak @taoeffect.com · 15/09/2026
Case closed. Worth reading the entire thread. Via: x.com/Hesamation/s...
000
Greg Slepak @taoeffect.com · 14/09/2026
Let China go large. We should do the opposite and go small.
000
Greg Slepak @taoeffect.com · 13/09/2026
In a surprise turnaround, @finkd stands out from these CEOs as being the more "honest" one (based on this search courtesy of Grok). Although all of us are still wondering when — if ever — that "open source" release of Muse Spark will happen.
000
Greg Slepak @taoeffect.com · 13/09/2026
Correct. @DarioAmodei is lying. @elonmusk is lying. @sama is lying. All of these people are either ignorant of how LLMs work, or are deceiving you.
100
Greg Slepak @taoeffect.com · 13/09/2026
Healthy response 👇
000
Greg Slepak @taoeffect.com · 12/09/2026
The "pacing the frontier" narrative is another psyop to combine 2 narratives: 1. "AI superintelligence singularity is here" narrative (false) 2. The "we need 'AI safety'" narrative (psyop to kill open source AI) Reality: labs are hitting the limits of what autocomplete can do
100
Greg Slepak @taoeffect.com · 11/09/2026
I was told the improvements would continue.
000
Greg Slepak @taoeffect.com · 11/09/2026
Still waiting for someone to crack our $10k bounty with one of these models. 🥱 Been waiting a long time. Find a flaw in our protocol, break into the chatroom, follow the instructions and it's yours. Database is right here: bounty.groupincome.org
000
Greg Slepak @taoeffect.com · 10/09/2026
Correct 👇
000
Greg Slepak @taoeffect.com · 08/09/2026
Some said I was making it up when I noted these models cannot do anything new. The implication being that anytime they claim otherwise they're taking credit for other people's work. x.com/hosseeb/stat...
000
Greg Slepak @taoeffect.com · 28/08/2026
. @AnthropicAI PR Marketing team: "Nobody will notice our watermark".
011
Greg Slepak @taoeffect.com · 26/08/2026
According to www.anthropic.com/news/claude-... it doesn't affect the length of responses, in which case it's just their training choices. H/t @_bgwoodruff
000
Greg Slepak @taoeffect.com · 24/08/2026
TIL pedophilia is legal in California and that @grok is broken. Via: x.com/JimmyJ4thewi...
000
Greg Slepak @taoeffect.com · 20/08/2026
What does it mean to "understand" something?
What does it mean to "understand" something?

Look at the word, it means you're standing under something.

You've placed yourself below a concept. You see it fully: its origins, the reasoning behind it, everything.

It is as if they came before the concept itself. Perhaps "beforestand" would be better.

Either way, when someone understands something, they can make authoritative statements about it, because it's as if they were the ones who created the concept themselves.

You cannot fight or argue with someone who truly understands something. This is why, "great minds think alike" is true. If they both understand the thing, they'll both agree.

Likewise, they will find it easy to make statements that seem surprising and yet are true about the concept to others who do not understand it.

They will be able to speak with authority on it, and they will be able to answer any and all questions about it without contradicting themselves.
000
Greg Slepak @taoeffect.com · 19/08/2026
And now @DanBilzerian seems to be moving on to POTUS election as if nothing happened. How does that make sense? You said he/they cheated. You're just moving on from one fake election to another election that you know is also fake? 🙃 x.com/ElectBilzeri...
000
Greg Slepak @taoeffect.com · 19/08/2026
Oh look fake democracy. If only there were some cryptographic way to ensure election integrity in a way that you can verify yourself… Kinda like we cover on the @okturtles.org Podcast…
020
Greg Slepak @taoeffect.com · 11/08/2026
. @finkd mogging US 🇺🇸 AI CEOs @DarioAmodei and @elonmusk. Shipping open weights while the others disfigure their models and have nervous breakdowns over CoTs leaking.
000
Greg Slepak @taoeffect.com · 11/08/2026
Could this be the reason for Claude's incomprehensible prose? 🤔
000
Greg Slepak @taoeffect.com · 23/07/2026
These are the best results I've seen yet for an A3B model. Congrats to the @KwaiAICoder team! > KAT-Coder-V2.5-Dev is now open-weight. Built on Qwen3.6-35B-A3B and post-trained with the proven KAT-V2.5 recipe huggingface.co/Kwaipilot/KA...
000
Greg Slepak @taoeffect.com · 10/07/2026
These are the default instructions and you can add your own when invoking it. It works great with my tasks skill: github.com/taoeffect/ta...
100
Greg Slepak @taoeffect.com · 02/07/2026
> @hashbreaker has been placed under moderation during this vote [4]. His posts require chair approval with up to a two business day delay. The stated reason is a copyright footnote.
000
Greg Slepak @taoeffect.com · 01/07/2026
Do you notice how there are no election related videos on YouTube? It's because it's not election season, and so the algorithm appears designed to artificially suppress all election related tech. We are learning so much about YouTube's algorithm just by covering this topic.
000
Greg Slepak @taoeffect.com · 30/06/2026
Congrats to x.com/youssofal_! I ran into issues last time I tried it, but now the latest version is indeed the fastest/best local coding experience on the Mac! mtplx.com
000
Greg Slepak @taoeffect.com · 29/06/2026
The #DigitalPrison preys on your sense of vanity. It sends you thousands of followers and likes and makes you think you matter. You must be willing to foresake all of that to escape the prison. You must be like @aantonop, willing to speak to an empty room!
100
Greg Slepak @taoeffect.com · 29/06/2026
crib.social/notice/B7pIK...
100
Greg Slepak @taoeffect.com · 29/06/2026
Hero. QT x.com/JOKAQARMY1/s... > A public service announcement. Bro said you need to drop them burgers and fries and stop eating at these fast food restaurants. Or you will be looking like him. 👀 🤣
000
Greg Slepak @taoeffect.com · 28/06/2026
I completely agree that "open source AI must win". Unfortunately the US's position translates this to "China must win", because 🇨🇳 has been leading in open source AI. Not a nice position for the US 🇺🇸 to put itself in. QT x.com/TheAhmadOsma... "Teamed up with NVIDIA to make Local AI The Default"
000
Greg Slepak @taoeffect.com · 24/06/2026
YouTube has completely shadow banned our recent okTurtles Podcast episode. 0 views via their algorithm in ~24 hours. That's why we share on loops.video and odysee.com The podcast is also available on most podcast platforms.
100
Greg Slepak @taoeffect.com · 23/06/2026
> What an absolute chad xcancel.com/TXMCtrades/s...
000
Greg Slepak @taoeffect.com · 21/06/2026
"Privacy for me but not for thee [and I'll send some dudes to break down your door and ransack your apartment if you try to violate it]" — Prime Minister of Denmark Mette Frederiksen Well, "if she has nothing to hide" what does she care? x.com/LarsAnders16...
010
Greg Slepak @taoeffect.com · 19/06/2026
We've updated the AI section in our contributing guidelines to require mandatory self-review from PR submitters who are submitting AI-assisted PRs.
### ❗️ Mandatory Self-Reviews on AI PRs

AI-assisted PRs are really PRs from an AI with human assistance. Therefore the human submitting the PR is expected to understand what they are submitting.

They cannot do this if they haven't reviewed the code the AI wrote. Therefore we **require** that the human submitting an AI-assisted PR review that PR.

We understand that Github's UI doesn't let you "review your own PRs", so here's what we expect:

- Go line-by-line through the PR. Post comments on anything that needs improving or fixing
- Have your agent use the `gh` command to fetch the oldest unresolved comment, and reply to it and/or perform the fix. Then push a commit with those changes.
  - Repeat this in a fresh session for each comment you left on the PR
- Finally: post a comment saying "Approved!" or something to that effect to let us know that you've (1) reviewed the entire PR yourself, and (2) fixed all the issues you found.
000
Greg Slepak @taoeffect.com · 17/06/2026
What are developers needed for in an age of AI like Mythos and Fable? Reviews. The code works but it stinks! Human reviews are still needed. I am still better than Fable and Mythos at reviews. Slower, yes, but significantly better.
100
Greg Slepak @taoeffect.com · 16/06/2026
New Dez Release! 0.6.1 adds support for #Venice internet search, web URL scraping, and X search (via Grok) Adds support for Z.ai's GLM-5.2 model. And disables the Venice system prompt so that you get the raw model. github.com/taoeffect/de...
020
Greg Slepak @taoeffect.com · 15/06/2026
China has won the AI race. Satya's vision of silo'd AI is spurred on by US regulatory action. In the short term, most people won't be able to make their own frontier AI. They will be forced to use China's. Iron Curtains wouldn't prevent it, just lead to Soviet collapse.
100
Greg Slepak @taoeffect.com · 14/06/2026
In the last few days, Chinese labs have given us MiniMaxM3, KimiK2.7 and now GLM5.2. Meanwhile US is censoring models. Reads like fiction.
010
Greg Slepak @taoeffect.com · 12/06/2026
Via: x.com/naomibrockwe...
000
Greg Slepak @taoeffect.com · 11/06/2026
> Thoughts on Fable from a friend. > > Builders beware. Via: x.com/adamislucky/...
Shows Fable sabotaging an ML project
000
Greg Slepak @taoeffect.com · 10/06/2026
Heads up that @AnthropicAI's Fable might be both more expensive and worse at reviews! news.ycombinator.com/item?id=4846...
000