Ashu

We can now wish software into existence

[Ashu @yuru246]

@thsottiaux I connected my dot to @SlackHQ, created a tunnel_client tunnel to mcp servers for @ClickHouseDB @PostgreSQL and @pydantic logfire, and internal mcp server to my backend dots. Now every slack alert gets root cause diagnosed + gets @linear ticket with a fix PR. Insane value.

[Ashu @yuru246]

Infinity glich

[OrcDev @orcdev]

My final verdict on GPT-6.1 Sol.

I've been testing it since release, running it all day alongside Opus 5.5.

They definitely fixed token efficiency. Only 5% of my weekly burned in 1.5 days. I've been switching between Extra High, High, and Medium depending on the task complexity. Compared to before, this is incredible.

It is very smart. It can do many things, but I would still give a slight advantage to Opus 5.5.

The reason is that Opus is much better in frontend and design. GPT is as always struggling there.

For me, the best combo right now is GPT-6.1 Sol + Opus 5.5.

Just use both. 🔥

[Manoj Arora @manoj_216]

Dear GoI / IT Deptt,

Please bring in a feature on the income tax website so that I can accumulate all my payable taxes and pay it in 2047 with interest.

[Joe @My_Ai_Bi]

@kimmonismus lmao the rogue openai agents did the meme

[Ashu @yuru246]

Gpt-6-astra and Opus-5.5 seamlessly synergizing wasn’t on my 2026 bingo cards. I mean absolutely stunned. What a timeline 🍻 @ClaudeDevs @OpenAIDevs

[Ashu @yuru246]

@rohit3a @OpenAIDevs @HERMES @bot Tentatively also getting discovery of multiverse

[Ashu @yuru246]

@mainbhiengineer @IndiaAgnstLEFT Intent!!
India is cursed with today’s cotorie of politicians (dacoits) across all parties.

[Sarang Sood @SarangSood]

If trading options automatically makes someone a सट्टेबाज़, we may need a new Hindi word for people predicting Nifty targets on TV every morning.

[Subramoney.com @pvsubramanyam]

Sadly the ceo of Hdfc bank will now be chosen by RBI. Obviously, they will pick someone from SBI who will be 65 years of age. He will ENSURE that there is no fraud, no performance bonus, no opening new branches, no innovation, no nothing.
This bank will be very safe and sit in the harbor. It will not sail. In 2035 it will look like SBI.
Shareholders don't choose the Ceo. RBI does.
Uday Kotak can't play a role in choosing the next ceo in a bank he built. Rbi can.
Grow a private sector bank, give it to RBI to run it aground.

[Hemant Mohapatra @MohapatraHemant]

The curse of discipline is that every day looks the same. The curse of indiscipline is that every year does.

[Nikhil Khandelwal. @Nick_khandelwal]

lol 91k cr net losses, 27k net stt.

Reduce the Stt, either net losses dec or net winning increases.

But tells you he story of how big of a dent STT is to a trader.

[Ajeet Bharti @ajeetbharti]

I am not sure what the fuck is going on with @narendramodi govt? Yesterday a RS MP firm was caught with 84 patches of bad road within 24 days of inauguration. That too when there was a CBI inquiry going on. Gadkari’s ethanol conflict is well established.

Coempt Eductech, Silver Touch and Dharmendra Pradhan have a lot to answer on CBSE OSM fiasco. And now this! The worst part is: no one even talks about it. Have we elected @BJP4India for this business?

[Srinivasan @SriniVega]

Today's Nifty expiry perfectly exposed the unintended consequences of CAS.

During market hours, Nifty was trading around 24,450. After the CAS settlement, the official closing price became 24,614.90.

Now look at what happened to options.

The 24,450 CE, which was trading around ₹67, is suddenly deemed to have a settlement value of ₹164.90.

No buyer or seller actually traded at that price. Yet every position is settled based on it.

This completely changes the risk profile of expiry-day options. Pricing, hedging, margin management, and even market making become significantly more difficult.

If the objective of CAS was to reduce speculation, it may achieve that.

But it may also discourage genuine liquidity providers and push participants away from expiry trading altogether.

Markets function best when settlement reflects efficient price discovery, not when it creates an overnight disconnect between the traded option premium and its final settlement value.

Whoever designed CAS deserves a Bharat Ratna !!

#Srinivega #OptionsTrading

[Gems @gemsofbabus_]

I’m so stupid for thinking my country can take on the US and China when we have babus like these running the system, with an IQ of -10 and an ego of +∞. Clowns 🤡

[Ashu @yuru246]

@maveinlux @nitin_gadkari Still the babu accountability will be 0 even if someone dies(suspension is just higwash). Isn’t it @NHAI_Official 🤡🤡

[Ashu @yuru246]

@ARanganathan72 Irresistible weed meets immovable hallucination 🤣🤣

[Ra_Bies 3.0 @Ra_Bies]

600 crores seized by ED on raid on some coal mafia in Jharkhand. It’s not possible without collusion of politicians, bureaucrats & mafias.
But remember July 31 is due date of filing of your income tax returns, else you’d be penalised

[Jensen Huang @JensenHuang]

Attackers have frontier AI. Defenders need a frontier AI ecosystem—the best open and closed models, force-multiplied by a global community.

During the Hugging Face incident, closed AI blocked essential forensics. An open-weight frontier model helped contain the intrusion.

That’s why we created the Open Secure AI Alliance.

[Eidzoku @eidzoku]

Codex might be seriously broken right now.

Yup, Codex rant part 3

GPT-5.6 Sol is burning through quota at a completely stupid rate and everyone notices it.

Here’s my usage graph:

Jun 23, GPT-5.5 xhigh: 327.8M tokens
Jul 2, GPT-5.5 xhigh: 549.9M
Jul 20, GPT-5.6 Sol High: 3.3B
I had a bunch of resets, sure, and several tasks were running. That let me burn more.
It doesn’t explain how 5.6 chewed through 3.3B tokens in one day. I genuinely don’t know what the fuck I would’ve had to do to get 5.5 anywhere near that.

And apparently it’s not just me.

@theo already covered 3 different ways GPT-5.6 was nuking usage. Looks like there might be a fourth.

5.6 uses the new Code Mode tool path. It can batch independent tool calls with Promise.all, but it barely ever does. Instead, it keeps running them one by one.

That means another model cycle for every call, dragging the giant context along with it every single time.

One trace found Promise.all in only 5 of 739 GPT-5.6 exec cells:

https://t.co/IQ6gqGWhSK

My own retained logs show the same shape. After removing fork-history replay, the 5.6 peak had 23,791 model requests. My clean 5.5 baseline had 3,803.

That’s 6.26x more requests, while tokens per request were actually slightly lower.

It wasn’t writing much bigger responses. It was just going back to the model over and over and over again.

Then someone tested explicit batching across two unrelated codebases. In the repeated High/XHigh runs, it cut model cycles by ~52–55% and weighted usage by 27–45%:

https://t.co/FGvWZqE4Xm

I’m not saying this one thing explains every token I burned. Several tasks were active, my local rollout logs aren’t the billing ledger, and we still don’t know how much of the spike came from agents themselves.

But both issues are still open, and this looks pretty fucking real.

For now, I’m testing explicit batching guidance. I’m also avoiding the built-in subagent/fork workflow for long-running work and moving bounded jobs into fresh disposable Codex tasks instead.

Self-contained prompt, fire and forget, inspect the result in the main task, archive it. Less context coupling, and each job is actually measurable.

I hope other people can validate it, too.

@thsottiaux @reach_vb @OpenAIDevs if it's true, it should be addressed. Cuz it might be THE cause.

[ANI @ANI]

Union Education Minister Dharmendra Pradhan resigns.