jingtao

I want wings

[jingtao @BLESSEDJINGTAO]

If you're frustrated with fills on the apps or want to experiment with multiwalleting you should get on basedbot. It'll help your execution noticeably.

https://t.co/pdjEnypPqr

[jingtao @BLESSEDJINGTAO]

@MomentumKevin I get the intent truly just frank is a legit dude and u guys would be friends

[jingtao @BLESSEDJINGTAO]

Frank beat the respect out of me, truly. I really didn't wanna give him any credit for a couple of years because I pigeonholed him as a dude who made it off making a NFT. In reality he's just very high aptitude and agency and has a nuclear motor. He's a born winner and a grinder, like you are.

[jingtao @BLESSEDJINGTAO]

I don't get the beef here. Frank's legitimately really talented on chain. He works harder than like anyone, builds his own tooling, takes his own ideas and tries hard. Maybe he was part of an op. Whatever the fuck. He's a good dude and you guys have more in common than you'd think. Nobody works harder than him.

[jingtao @BLESSEDJINGTAO]

@orangie @KothCap That's hype. I think I'm going to get in the lab to try to make some new research tooling for agent stuff this week. Have some funky ideas w/social graph stuff and on chain I wanna experiment on. It's the best time in history to try to make and try new ideas. Gl out there matey.

[jingtao @BLESSEDJINGTAO]

@fomocapdao Hmmmm

[jingtao @BLESSEDJINGTAO]

@orangie @KothCap Never seen #3 before

[👀 @UniswapVillain]

Its extremely rare there's a crypto token that genuinely is likely to get acquired and when doing so its holders also get that exit payout.

Simple math.

30M raise valuation x 40% (token share of equity) = 12M valuation for token

$WSOLP and chill.

[Jubal Hardin @Jubal_Hardin]

If you think that you were surprised, imagine how surprised the observer in the next layer up was watching you create your sim the exact same way they created theirs.

Now they have to cope with the probability that their reality is also just a projection on a 2D screen boundary emerging from organized quantum data.

Unreal all the way down.

Things just get weirder from now on. Very quickly.

e/acc ai/acc

[jingtao @BLESSEDJINGTAO]

Because I was scared of my own shadow and didn't trust myself after a while of running bad. Just need to stick to what I know to be true and remember what made me elite at some point in the first place.

[Viltrum @VILTRUMsol]

@PUNISHEDJINGTAO Ty for the 3x but I am curious why you sold

Is it a scam or is your original thesis after DD in tact

[jingtao @BLESSEDJINGTAO]

Hope you guys crush

[jingtao @BLESSEDJINGTAO]

Took the 10x on rim gg

[jingtao @BLESSEDJINGTAO]

It's been six years I been doing it wrong this whole time! Doesn't matter how good I am if I'm a net slaughterer

I'm choosing to believe WITH now

[jingtao @BLESSEDJINGTAO]

I'm leaving this entire pvp slaughterer persona behind

We will win together

I'm holding hands

I'm kissing babies

I'm singing kumbaya

We can simply nuke on retail

Let's work

[jingtao @BLESSEDJINGTAO]

I'm leaving this entire pvp slaughterer persona behind

We will win together

I'm holding hands

I'm kissing babies

I'm singing kumbaya

We can simply nuke on retail

Let's work

[jingtao @BLESSEDJINGTAO]

Happy you guys won let's keep cooking more and more spots incoming

Time to lock in

[jingtao @BLESSEDJINGTAO]

My dd on pump app

[jingtao @BLESSEDJINGTAO]

WE'RE ALL GONNA MAKE IT......HOOOOOOOOOOOOO

[jingtao @BLESSEDJINGTAO]

NO MORE PUNISHED JINGTAO

WE ARE RUNNING GOOD NOW

[jingtao @BLESSEDJINGTAO]

I already sold it enjoy

[Eric Cryptoman @EricCryptoman]

Still doing dd on $zzz but if it’s real should be a banger

0x7dbf38976f6d3b9c529e7d9484a71898b409ee6a

Anyone else got any info to share? From what I’ve seen so far it looks real - still not confirmed by @GMX_IO but if they do I imagine a 25m+ candle

Those guys are OG shippers

[Matt Shumer @mattshumer_]

A lot of people are asking how I pulled off these super long-horizon builds with Astra.

Astra is extremely powerful, but by default it struggled with a task this difficult. I tested a bunch of approaches to get past this, and the one I landed on is something I'm calling the Manager Loop.

It's basically a couple of tricks we used to use with much less capable models a couple of years ago, with a few new ideas layered on top. Turns out that when you put those together and apply them to Astra, its ability to do extremely difficult long-horizon tasks goes up dramatically.

Here's how it works:

1. Launch an agent (I'm calling this one the "manager"). Chat with it about what you want to get done, and have it build a massive checklist of to-dos, then break that checklist into phases.

2. The manager then spawns a second Codex agent in a separate thread (the "implementer"). The two agents can message each other.

3. Put the manager in /goal mode, and tell it to run each phase on the implementer in /goal mode.

4. The manager messages the implementer: "/goal Complete phase one completely, extremely well." The implementer doesn't stop until that phase is done, then messages the manager back. The manager tells it to start phase two. They repeat until every phase is finished, completely autonomously.

Why I think this works: over a long-horizon task, Astra tends to asymptote. It gets way further than previous models, but at a certain point it kind of just stops improving against the goal as quickly as it did before. It gets stuck in the minutiae, focusing way too much on small details, and overall progress stalls. The Manager Loop forces it to work piecemeal, one phase at a time. It's essentially how a human would steer a model, except the model is doing the steering for me.

That's actually how this started. I was having the model write the checklist and break it into phases, and then I was doing the manager's job by hand. At some point I thought, "Wait, why can't I just get a separate AI to do this?" That's what unlocked full autonomy, which is super useful.

A wording detail that seemed to matter: I ask for each phase to be done "extremely well," not "perfectly." Maybe I'm reading too much into it, but asking for "perfect" sent the model right back into the minutiae. "Extremely well" implies it's allowed to move on once it's good enough, and that worked better in my testing.

One more trick that I think helps (this one is more of a hunch, but it was useful for me): have the implementer build a simple HTML page with the full checklist on it. The implementer checks boxes off as it goes and updates a counter, and the page has a chart of # of boxes ticked over time.

Obviously the boxes aren't all equal, but it forces the model to notice things like "I haven't made progress in a while, time to move on." You can even put this in the prompt directly, like: "if you haven't ticked a box in X amount of time, move on". That helps a lot.

I also ran 96 sub-agents at a time. You can change this in your Codex config (or just ask Codex to change it).

This got me far better long-horizon performance than anything else I tried. I'll be sharing more in the coming days!

[jingtao @BLESSEDJINGTAO]

My dd on pump app

[jingtao @BLESSEDJINGTAO]

Forgot how much knowledge gap there is from 2024 entrants to before- the GMX deployer just launched something and it's not auto 10m