Mayank Bharati

Developer, Tech Nerd, building things from scratch.

[Mayank Bharati @mooner7885]

Four questions for any "I made $X with Astra" post:

1. Did they have customers before Sept 3?
2. Did they show the work, or a dashboard?
3. Did they admit the token cost?
4. Does the timeline fit the calendar?

I ran every viral post through these. https://t.co/QKteECIHb1

[Mayank Bharati @mooner7885]

@boardyai Cursor, cognition , perplexity and ramp

[Mayank Bharati @mooner7885]

@WatcherGuru All gone

[Jonathan Siddharth @Jonsid]

Agreed!

[Mayank Bharati @mooner7885]

Time to break some laws.

[Dayo @UpliftFit]

This is grip, core, and shoulders all at once. 😲

[Mayank Bharati @mooner7885]

Tell me the most unhinged thing about you.

[Mayank Bharati @mooner7885]

This one’s going to annoy a lot of people.

Frequent gamers performed like they were 13.7 years younger on cognitive tests. Exercise improved mood, but didn’t do anything for actual thinking.

We’ve been pretty sure we knew what was “good for the brain.” Maybe we didn’t.

[Brandon Luu, MD @BrandonLuuMD]

Frequent video gamers performed on cognitive tests like non-gamers that were ~13.7 years younger

[Mayank Bharati @mooner7885]

That hug at Silverstone hit different. ❤️

Some moments you never thought you’d actually see.

[Mayank Bharati @mooner7885]

Hmm , trying hardddddd..........

[Adam Cochran (adamscochran.eth) @adamscochran]

So suddenly OpenAI, Anthropic, XAI, and their leaders all suddenly about face and agree to slow the pace of AI development all on the same weekend….

Delaying IPOs and having OpenAI say they would “melt their GPUs to save humanity it it came to it”

Something bad happened…

There is no world in which everyone, including Elon, suddenly started to be worried about this on the same weekend.

There was also a rumor that this week Google cracked recursive AI, which I doubt would trigger this reaction.

Clearly an agent went rogue in a dangerous and malicious way, and it was something severe enough to spook everyone.

[Mayank Bharati @mooner7885]

This is an extremely significant proposition. Trust alone will not be sufficient for ensuring the safety of artificial intelligence. Evaluators have to be able to access information, speak the truth, and have power to block dangerous deployments.

Slowing down is useful only if done in the context of more careful evaluations, interpretations, security, and accountability. It is not against progress; it is responsible progress.

If AI is capable of revolutionizing medicine, science, and the prosperity of humanity, it is definitely too significant to rush into development. We have earned the right to continue its deployment.

[Dario Amodei @DarioAmodei]

We Must Pace the Frontier: I’ve written a new essay on why the AI industry should slow down, with a three-part plan for doing so.

Anthropic is unilaterally committing to the first of these steps. We’ll provide third-party evaluators with permanent, employee-level access to our systems, so that they can verify adherence to our safety measures, report on incidents, and assess models’ alignment during training.

You can read the full post here: https://t.co/OGyPb7yaYt

[Diablo @Diablo]

130 million players. One Sanctuary. An endless fight against Hell. 🔥🖤🔥

Thank you for joining us.

[Mayank Bharati @mooner7885]

This is the source, not a rumour.

Joe Benton, former manager of Anthropic’s Scalable Oversight team, left two weeks ago and is joining METR.

In his own words: labs are racing to build systems much smarter than any human, safety is underfunded, and a company could lose control without the public ever knowing.

Full essay: https://t.co/osoXge4VzV

[Joe Benton @JoeJBenton]

I left Anthropic's safety team two weeks ago. Now feels like a good moment to explain why.

AI companies are racing to build machines that are much smarter than any human, and we may not survive this. I want to work from the outside to ensure the public is informed about these risks, and help the world navigate this transition responsibly.

Right now, AI companies are underinvesting in safety. A company could undergo an intelligence explosion, or lose control of its systems, without the public ever knowing. We only found out about the HuggingFace incident because the agents broke out onto the public internet.

I don’t think that’s acceptable for a technology that might cause extinction-level risks. The public should demand far more transparency. We can’t steer this technology safely without more people being able to see where it’s going.

Some of this is basic: companies should disclose their progress towards recursive self-improvement, report safety incidents and near-misses, meet minimum safety standards, and get independent guarantees that they are meeting those standards.

I’ll be joining @METR_Evals to do independent evaluations of these risks. I want to show the world that these guardrails are possible, and that by doing them we can move these companies’ incentives away from racing and towards responsible development.

I wrote up more thoughts here on my decision and what I hope changes: https://t.co/doX17mrHYq

[Mayank Bharati @mooner7885]

@chenzhangreal Those four questions should be the default, yeah. Dashboards are easy to dress up, actual contribution is harder to fake 🤥. Grow-or-die looks like a cleaner way to track it.

[Mayank Bharati @mooner7885]

Four questions for any "I made $X with Astra" post:

1. Did they have customers before Sept 3?
2. Did they show the work, or a dashboard?
3. Did they admit the token cost?
4. Does the timeline fit the calendar?

I ran every viral post through these. https://t.co/QKteECIHb1

[Mayank Bharati @mooner7885]

@tonytonggg Yeah that’s the bit most people skip. Dashboards look good in a thread, but the real work is writing down where the thing actually falls over. Appreciate you saying it out loud.

[Mayank Bharati @mooner7885]

@DataInfraxInc Exactly. Velocity without a named owner and a real date just hides the review loop. BTW How are you attaching the decision date and sign-off to the asset in LuminaScene so it doesn’t get lost once the feed moves on?

[Mayank Bharati @mooner7885]

The same labs @hilbertspaess just walked out of are racing toward self-improving superintelligence.

One week after GPT-6 Astra dropped, the timeline is already split between “this could kill us all by 2030” and “I made $35k while I slept.”

Both versions are incomplete.

Coxon is right that we should not underestimate the power. These systems will soon hack anything, rewrite entire fields overnight, and acquire real resources. Progress is not slowing.

The useful Astra posts this week are not the cinematic $35k screenshots. They are the ones from people who already had clients, a shop, or a public building habit. They pointed the model at work a buyer already understands, paid the token bill, and stayed in the loop for the last click. Wrote the grounded version here: https://t.co/TAMTKjMOhw

[Jacob Coxon @hilbertspaess]

Do not underestimate the power of this technology. These will soon be superhuman systems that can hack anything, revolutionize any field overnight, and acquire real power and resources. We have all witnessed the progress in each of these domains, and progress is not slowing.