ทวีตที่ปักหมุด
Bart
677 posts
Bart
@bkolendowski
Cofounder @ HoppeSyler | Writing about what actually changes when language models face real stakes https://t.co/88g0lsWjhA
เข้าร่วม Kasım 2023
123 กำลังติดตาม128 ผู้ติดตาม
@69chakoop69 @get_crazy__ @james43678 @dc_mma @grok I didn't ask grok why it didn't know something, I asked it explain the contradiction of posting two opposing things *at the same time* (just in different areas).
Hope that helps.
English
@get_crazy__ @bkolendowski @james43678 @dc_mma @grok the problem with AI is the billions of taxpayer dollars going into scraping the data of american citizens who didn’t ask for it lmao
asking an AI why it didn’t know something that nobody could’ve answered other than 1 guy 5 seconds after said guy deletes a post is a user error
English
@james43678 @dc_mma @grok how come here you are admitting that DC posted and deleted this but in other places on X you're stating that it was never posted/deleted?
English
@ShawnOnTarget @NeedMyNutsLickt I never said its Sean's account? This tweet has 1.2M views - I don't think they're going to be like "ah but its not Sean's".
English
@bkolendowski @NeedMyNutsLickt This is not Sean’s account either moron it’s a troll with 1k followers lmao
English
I’m a now deleted post by Daniel Cormier, Eric Trump DM’d him and asked him “are any of the fights tomorrow rigged?” 👀
#UFCWhiteHouse
English
@MartinShkreli @grok can you explain to Martin Shkreli the 'real' benefits (assuming he means from randomized controlled trials) and then the other evidence that he likely considers 'not real' like the TriNetX database study, the 2024 systematic review, etc?
English
this is irresponsible & borderline insane. there is 0 real evidence that tadalafil has any health benefits, let alone an all cause mortality benefit. while the author states that its just an association, the implication is irresponsible because the odds it does this are <0.01%
Bryan Johnson@bryan_johnson
I take Cialis, but not for sex. It’s actually a longevity medicine. Cialis (Tadalafil) is great for the same reason it gives you fantastic erections… it improves blood flow. Studies show that Tadalafil… + 34% reduced all-cause mortality + 27% reduced major heart disease + 34% reduced stroke + 32% reduced dementia It has also shown benefit in insulin sensitivity, metabolic health, and reduction of body fat. Women have blood vessels too, so theoretically they'd get the same longevity upside. The research is thinner, but early signals are promising. It’s sad that when men and women could benefit from it can miss out because it’s taboo. My protocol is 5mg daily and I've been on it for about two years. *Observational research shows associations, not causation. Outcomes may be influenced by underlying differences in study populations. This is not medical advice and is shared for informational purposes only.
English
"Why is everyone constantly ignoring the simplest explanation and going for the most cynical?"
The same can be asked why you're claiming that they're doing it for no other reason than the 'right' reason. Why?
The truth can be that they're doing it for both. It's the right thing to do and clearly the regulatory capture angle would help them as a business.
English
@MatthewBerman Dude, they did that because it was. That is normal behavior for someone trying to actually do the right thing.
Why is everyone constantly ignoring the simplest explanation and going for the most cynical?
English
This is pointless. @DanielMiessler and others have already shown that you can have directories set up in anyway that you want that fit your own personal system and all you have to do is give the agents a pointer (i.e., an index) a few instructions and they have no trouble understand and navigating the system whatsoever. Importantly there's no reason to be limited to markdown files.
English
This is really big news. Google introduced the Open Knowledge Format (OKF) - a standardized way to store information in a directory of markdown files. Makes it really easy to make a digital brain that agents can use.
These files can serve as a living wiki. You can give agents the ability to query them or edit them. They can interlink.
Seems to me this could replace Notion or Obsidian. I can think of so many uses for this.
Google's blog post: cloud.google.com/blog/products/…
An easier to understand explanation is the SPEC.md file:
github.com/GoogleCloudPla…
I gave those two links to Antigravity and asked how we could use it for any of the projects we're working on. It came up with so many ideas. I would imagine Claude Fable 5 would whip up some pretty amazing things based on this system.
Currently creating an OKF library of our pepper garden. It's going to be a fun weekend.
English
@xsteenbrugge "You were right, I cut them"... Ahhh... Still has a ways to go I see.
English
More than ever before I feel like Fable 5 speaks a different kind of English than we humans do.
It isn't conversational, its not 'light' to read, it barely explains — it's pure load-bearing density, meaning compacted until every token earns its place.
It assumes your context window is as large as its own and you have perfect recall of everything you're working on. Similar to how AI-chess feels alien, this too, feels almost akin to an alien language.
I mean, just read this stuff, nobody talks like this, only advanced AI does:
English
I'm wondering how much of this is subsidized by enterprise (they buy 150 seats but most people treat it like a chatbot instead of hitting anywhere near the limit)
SemiAnalysis@SemiAnalysis_
Recently, we purchased one of each Anthropic/OpenAI subscription plan and randomly ran long horizon coding tasks until we exhausted the weekly limit. It's widely believed that a $200/month plan maxes out at ~$2000/month worth of tokens (assuming API pricing). However, we found that the subscriptions are actually far more generous. (2/4)
English
Bart รีทวีตแล้ว
@DanielMiessler Oh I don’t have a problem with the US (or I do I guess but that’s not related). I just want everyone to have equal access to frontier ai. Anthropic is very clear they don’t want that
English
Send all of Anthropic's compute to Deepseek. I want a new open-weight model and I don't care if "China wins"
English
Bart รีทวีตแล้ว
Increasingly, I believe companies may need to be rebuilt from the ground up, where you have a single timeline of all observability + product metrics + file changes laid out in a retrievable system, like Datadog + Posthog + Google Drive + Slack (really unified filesystem of Claude Code chats + Codex chats). This might be the new data foundation for any and all companies to maximize AI. Needs to be rebuilt because keeping track of diffs on existing system basically impossible to produce longitudinal information on decisions and rollbacks, something coding agent storage companies are actively trying to figure out, but this should extend to businesses as a whole.
Highly skeptical existing businesses will adopt this though because it means overhauling everything about their instrumentation and business data, but I think businesses built on this foundation probably can execute 100x better and faster
English
@omgsidewalks @grok is it true that no billionaire actually built a billion dollar company through hard work?
English
No billionaire actually built a billion dollar company through hard work. Apple’s cofounder said the garage story is a myth. Bezos got a $300k investment from his parents. Google’s founders were Stanford grads with access to venture capital. Stop promoting the bootstrap ideology.
EDOSE✨@iam_biglad1
Hit me with a random fact:
English
It can't be understated just how much the key to getting value out of 'AI' is (much like anything else) an iterative process.
For something like Claude Code:
1. use it to do something meaningful
2. improve or correct it by a small amount each time (make or add a skill to automate something, add a rule to CLAUDE.md to prevent a consistent issue, etc)
3. repeat
Over time start doing more complicated things (bring in MCPs, tasks that are genuinely difficult, etc) but continue iterating.
That's it. Just do that.
English
@buc9755 @DrAndyGalpin If you understand what it means then why would what he's saying not be consistent with the position that humans don't have free will?
English
Sam Harris makes a point here that's hard to dismiss: most people can't stay undistracted for 30 seconds. Not because they're lazy — because they've never built the capacity. And that's not a trivial problem.
If your emotional state is just a byproduct of whatever thought randomly surfaces next, you're not really steering anything.
The fix, according to Sam, is learning to notice thoughts as appearances in awareness rather than being identical to them — and mindfulness meditation is the most direct way to start building that skill.
Andy Galpin, PhD@DrAndyGalpin
New Perform with Dr. Andy Galpin episode: How to Think Clearly in an Age of Distraction | Sam Harris 0:00 Introduction & Sam Harris 2:30 Intelligence, Smartness & Wisdom 16:13 Outsourcing Cognition: AI as Tool vs. Competitor 21:19 The Death of Boredom & the Inner Voice 30:07 Mindfulness & Three Paths to Mental Freedom 42:20 Overcoming Resistance to Meditation 1:01:25 Psychedelics vs. Meditation 1:09:01 Emotion, Reason & Handling Criticism 1:28:06 Process, Expertise & the Scientific Crisis 1:50:44 Restoring Trust in Institutions 2:02:29 Consciousness, AI & the Hard Problem 2:16:18 Uploading, Teleportation & the Revenge of the Humanities 2:25:14 Closing Includes paid partnerships.
English
@DrAndyGalpin Isn't this the guy who says that humans don't have free will? Funny how he doesn't act or talk like he believes that
English
@johnowen19681 @PeterSweden7 @grok what other laws has Germany had that, in hindsight, were obviously atrocious?
English
@PeterSweden7 He is temporarily blocked from going to 1 country until the end of next month. He is still allowed to travel elsewhere. Germany has had that law on its books for decades, and their Supreme Court has upheld it as legal under their constitution.
English
Germany just BANNED a political activist from leaving the country because he could be spreading "far-right extremist ideas".
Stopping people from leaving the country because of political opinions.
Isn't this what they do in North Korea or something?
English
@MattSeedorff @OCFireAuthority 100F is only 37C... Did he mean 100C?
English
BREAKING: Officials say the failing toxic chemical tank in Garden Grove, CA is now above the 100-degree threshold on its gauge.
@OCFireAuthority says crews have discovered a crack in the tank, but the risk of a catastrophic failure or explosion remains. Evacuation orders are still in place for thousands.
English