mustaphah

I don't like Ed, but it seems that Luu spends far more attention documenting failed predictions than checking other predictions that might have been correct.

Haidt, in his great book "The Righteous Mind," has been arguing that reasoning evolved not to discover truth but to win arguments. There's a lot of scientific research backing his idea.

Haidt's metaphor is the rider and the elephant: the elephant (intuition) leans, and the rider (reasoning) invents the justification afterward and then defends it like a lawyer, not a truth-seeker.

Intelligence doesn't fix this - it just makes people better at coming up with hard-to-defeat arguments; that explains why smart people disagree all the time.

You can probably catch a big pie of those with simple heuristics to flag suspicious repos for expensive review (human- or AI-based). I did that with public account & repo data, and I believe they can do much more given the amount of private data they have access to.

I'm talking about 10s of repos flagged in a few hours. I don't think the volume would be that big for an expensive review.

Well, my trend detection logic rewards recent stars more than older ones [1]. Recency is an important factor for many custom and public tools that track GitHub trends. I think the bad guys intentionally recreate repos - I actually noticed that.

That being said, they do take action if you report the repo. So I'm guessing good users are doing the heavy lifting here with reporting. I don't believe GitHub is taking enough proactive measures, or maybe they do, but it's not working well, obviously.

https://hadid.dev/posts/github-trends/#growth-based-approach

This is just one flavour of abuse. GitHub does NOT give a shit about the scale of the malware problem.

I've seen so many forms of malware repos working on a GitHub trends newsletter [1], mostly about crypto, NFTs, KMS, and similar stuff.

In the first runs of the project, I was so surprised by tens of malware repos that looked like trending repos. A lot of them share some common traits that made filtering feasible:

- Made by a fresh GitHub user - many created in the past few days.

- The average creation date of Stargazers accounts is very close to the repo creation date. If you take the mean time diff, those bad repos get exposed.

I reported 10s of malware repos, but then I gave up as I felt GitHub was not really doing enough to fight back. I was like... these guys don't seem to care, why should I?

God knows how many people have been abused by these malware repos on GitHub.

---

[1] https://github.com/mhadidg/gh-trends

You want my email voluntarily for the whole purpose of telling me "Hey, Fable is back"?

Everyone would be screaming the moment that happens. No, Thanks!

I finished The Lean Startup a few days ago, and I really felt the power of the ideas you've shared there coming from a heavily technical background; incredible work.

It's already been 14+ years since you wrote the book; I wonder if a second edition is something you have in mind, or at least on your consideration list

You would probably be better off reading learnxinyminutes.com/zig/

I have a strong feeling the whole thing is distracting the people - the real question is, distracting from what?

Sam Altman and other big figures tend to shape their narratives around their personal and organizational interests. When people were skeptical, they pushed hard into the "God-like AI" narrative. Now that safety concerns are growing and their growth plans are in danger, they're pushing back against what they used to advocate.

Even if they genuinely believe what they’re saying, their perspective is still fundamentally biased and should always be taken with a healthy grain of salt.

I hate to say it, but I'm becoming less and less interested in structured content, and more interested in disorganized, messy content over time. I don't like the thought of how this may end up in a few years for me.

"AI bubble" in the title, count me in.

Tokenmaxxing is so dumb. You should never show your team how exactly you're measuring their performance; people will optimize for the metric, not the actual performance.

Classic Goodhart’s Law: when a measure becomes a target, it ceases to be a good measure.

Feels like they are debating internally whether to cut people or AI spending. Very healthy debate. Let's hope they spare people.

Can't locate the link to the actual speech

> I'm not convinced he's all that smart

We probably disagree on the meaning of "smartness"

Couldn't tolerate the content; it's too structured; I'm open to something more chaotic.

Extreme smartness has its own failure modes

The strongest evidence against Musk was Musk. His own 2017 emails supporting for-profit chats made the "betrayal" narrative very hard to sell.

Maybe this explain it:

https://www.psypost.org/scientists-just-discovered-that-a-hi...

That's a fair clarification.

I never said we have sufficient evidence to act. But "too good to be true" + "singular paper" together can become an unfalsifiable dismissal - by that logic, every important result looks suspicious before it replicates. The interesting question is what priors should update our confidence here.

Stanford/Arc Institute and published in Nature + mechanistic grounding + prior research on gut-brain axis gives me way more confidence than average, but you're right, that's not nearly enough for most, but quite sufficient for me, and surely others with informed priors or a strong motive.

I guess "too good to be true" is not a decent argument to convince a rational mind

> After you read a lot of those you learn not to take individual papers seriously

Can't disagree, but keep in mind that almost all meds are tested first in mice/animal models before human trials verify the effects.

I guess it's because most major disorders and diseases have so many pathways at play that figuring out which one's actually causing the problem at the individual level is just too tricky.

The other thing concerns how potent the effect is to be therapeutic. In many cases, the effect is just marginal to be meaningful.

Chili could work [1], but not too much of it [2]

[1] https://pubmed.ncbi.nlm.nih.gov/27079706/

[2] https://pubmed.ncbi.nlm.nih.gov/31137805/

Yeah, it's a mouse study, but there are tons of human studies backing the whole gut-brain connection. There are even a bunch of books on it [1][2].

What's really cool is that the paper used low-dose capsaicin (just 5 μg/kg injected), and it completely restored hippocampal FOS activity and memory in older mice. Basically, that's the same stuff you get in cayenne pepper supplements - pretty easy to get your hands on.

[1] https://www.goodreads.com/book/show/28837738-the-mind-gut-co...

[2] https://www.goodreads.com/book/show/35210457-the-psychobioti...

Well, it's fixed now. Apparently, Claude Code was still up and running!

I'd bet it's Cloudflare's fault

Even worse, I got contacted through YC Jobs (workatastartup.com) with a message that was basically: "Star, fork, and submit PRs to our open-source repo and we'll review you for a contract."

I immediately realize it's engagement farming + free labor. I said "No thanks."

Got this reply: "(...) I'm looking forward to reviewing your PRs. Feel free to share me any of your questions. (...)"

Apparently, no one read my reply - not even AI. They are automating this shit. It's sad that many fall for it (check their Github repo)

---

Company: Aden (W20)

Contact: Vincent Jiang, Founder

Github: https://github.com/aden-hive/hive

Worth noting: emails from .online domain (and many other TLDs [1]) are also way more likely to end up in the spam folder.

https://www.spamhaus.org/reputation-statistics/gtlds/malware...