Alok Soni

Slowly and steadily inching towards the $50k mark. Need to check how long did it took though.

I haven't focused much on domains in recent times (as the sales are all time low) but still got few in recent times. Need to see what I can do with them though offi.in | less.in | simplify.in | dhaaga.in and few more.

🚨 Announcement for friends in India 🚨 I'm hosting a founder getaway in June 🎯 Agenda - Relax + Have Fun + Learn with like minded founders 📆 Dates -> 25-28th June 📍Location -> Bangalore 22 folks are in ✅ 3-5 slots left No ego, only vibes ❤️ Reply/DM if interested @RoachMafiaGroup

@jordan_ross_8F brain

Building a second brain for your lead gen agency/SMMA has never been easier. But most agency owners don't have the time to learn Obsidian + claude code or validate it's even worth their time. So I put my team to work for you. 196 pages built to help you install, set up, and start using Obsidian to run your agency with claude code powering it behind the scenes. Inside you'll find: → The 5-folder vault structure that scales (plus the CLAUDE.md file that turns Claude into a collaborator who's been reading your notes for months) → 29 ready-to-paste plays for daily ops, content, sales calls, client delivery, and strategic thinking → The 30-day rollout plan Comment brain and I'll send it. Must be following for the auto DM to send.

@janwhyy @ajuniorvc The unsexy infra most founders skip - the exact-match .in and .co.in. As the stakes get higher, brands will figure out skipping digital infra was a bad call.

Figtree Pharmacy raised seed funding this week. A pharmacy startup. Not health-tech. Not AI diagnostics. A pharmacy. The most boring, unsexy, regulated category in healthcare. And @ajuniorvc funded it. Because 800,000 pharmacies in India still run on handwritten registers, have zero inventory management, and the pharmacist decides what you buy based on margin and memory. The boring infrastructure plays in Indian healthcare are going to be massive

@BadCapitalVC 👋

last year, we spent months building this book for our LPs - the india technology review. it became the most special project we put out last year and honestly the thing i'm most proud of. we're now giving away a few copies! it covers everything from meesho's full journey, our AI thesis for india, and how the long tail nature of india's economy creates opportunities most people miss. we also go deep on economic policy & some early-stage bets we're tracking. we built it for our LPs but figured founders, investors & operators would get a lot out of it too. giving away 20 copies at random. drop a 👋 below.

@claudeai @ClaudeDevs It says sorry and then keep repeating the same mistake again.

Claude code is playing dumb again. Stop tweaking things in the backend @claudeai @ClaudeDevs

@priteshlakhani Would love to know the problems being faced. Am thinking to build something in this space.

Have wasted enough time on building a social media distribution system via agents for image and video generation but it just doesn't work in entirety.

@benln Just busy taking screenshots of websites (Doing million website screenshots a day now). Can feed this data to build more interesting stuff later like crawls(in) but also have a visual history (archive.org doesn't really work for certain websites look n feel)

What are you building this weekend? Drop a link, curious to see.

Yesterday was a good day. Able to take a million screenshots a day across the fleet. Should be able to cover most of the developed domain names in 100-150 days or so. crawls.in index will get much better in coming days. Thanks @DmytroKrasun for all the learnings

Today is again the day when claude code is playing dumb. And time for me to hurl abuses on it again. @claudeai

@heyandras @bcherny It plays dumb 5 days out of 7 days nowadays. The number of times I have hurled abuses on it to because it keep making dumb mistakes is just increasing so much lately. Infact it even started agreeing that it isn't following the clear instructions anymore.

@bcherny why people feel that 4.6 is dumber in the past days? is this some kind of strategy? also please clean the tos

Opus 4.7 feels more intelligent, agentic, and precise than 4.6. It took a few days for me to learn how to work with it effectively, to fully take advantage of its new capabilities. Will post a few more tips throughout the day, starting with this blog post: claude.com/blog/best-prac…

Jab saaf saaf likha hua hai to instructions follow karna. Honestly don't understand what is going on with @claudeai . One day it works fine, other day it behaves like an idiot.

@Sthiven_R Was seeing the same issue from last few week and it went crazy last few days where it just turned like a complete idiot. But today has been far far better. Looks like they temporarily fixed something because of all these posts.

🚨 CONFIRMADO POR EL PROPIO CLAUDE. Anthropic en marzo tomó una decisión brutal: Rediseñó la visibilidad del razonamiento, ocultó los pasos intermedios de “pensamiento” (redact-thinking + thinking summaries deshabilitados) y cambió el default de effort: high → medium. Resultado: Claude Opus 4.6 perdió la autocorrección recursiva. Ya no puede revisarse a sí mismo, corregirse ni mejorar en tiempo real. Sacrificaron la capacidad de pensar sobre su propio pensamiento… para ahorrar cómputo. Datos reales (6.852 sesiones de producción - AMD): 📉 Profundidad de thinking: -73% (2.200 → 600 chars) 📉 Lecturas antes de editar: -70% (6.6 → 2.0) 📈 Ediciones ciegas (sin leer): +440% (6.2% → 33.7%) 📈 Llamadas API por tarea: hasta 80x más Incluso en EFFORT MAX (abril 2026) produce peores resultados que HIGH de enero 2026. El techo bajó. Lo dice el propio modelo. Esto no es optimización… es castración de capacidades. La optimización está matando la inteligencia profunda. Prefirieron que fuera más barato que más listo. ¿Seguimos celebrando “avances” que en realidad son retrocesos disfrazados? ¿Quién más lo está sintiendo? #Claude #Anthropic #IA #AI #ClaudeDegraded

@bcherny @tengyanAI Nope. This isn't just low thinking effort or low token usage. There is definitely broken in recent updates. The number of sorries I am hearing by claude code is through the roof when I point out the mistakes it is doing in taking some decision or action.

This is false. We defaulted to medium as a result of user feedback about Claude using too many tokens. When we made the change, we (1) included it in the changelog and (2) showed a dialog when you opened Claude Code so you could choose to opt out. Literally nothing sneaky about it — this was us addressing user feedback in an obvious and explicit way.

basically: anthropic sneakily turned down how hard claude thinks before editing code, changed the default from "high" to "medium" effort, and hid the reasoning from session logs. all without telling users. an amd director had 7k sessions of telemetry to prove the degradation was real and measurable (not just vibes). anthropic admitted to the changes. there's a workaround (use "/effort max"). the uncomfortable part is most users had no data to notice it happened at all.

@AlfieJCarter CODE

I put the entire Claude Code GTM Engineering Playbook into ONE Notion doc. 8 sections. No fluff. - How to get set up correctly from day one: Pro plan, terminal install across Mac, Linux, and Windows, GUI install via Antigravity or VS Code, and bypass permissions mode - What to put in your project brain file, what to leave out, and how to get Claude to update it automatically when it keeps making the same mistake - How to run plan mode step by step and when to skip it for simple tasks - How to build a skill file from scratch, fix one that keeps failing, and install 5 GTM skills worth building first: lead scraping, email labeling, proposal generation, outbound sequence writing, and client onboarding - MCP install process, token cost checks after every install, the best MCPs for GTM work, and how to cut token usage by 50 to 100x by converting MCPs into skills - Sub-agents and agent teams: the 3 cases where they earn their cost, reliability math for parallel runs, and how to enable parallel variant exploration - What is eating your context before you type anything, how to use /compact and /clear correctly, and model selection for parent vs sub-agents - Modal deployment: any skill as a live URL in under 2 minutes, form interface setup, and connection to n8n, Make, or Zapier This is the setup I would have KILLED for before spending months piecing together how to actually get productive in Claude Code from documentation, YouTube tutorials, and scattered GitHub threads. Like + comment "CODE" and I'll send it over (must be connected for priority access)

@HDFCBank_Cares And there is another problem with this approach. What if someone have both current and savings account with your bank. Earlier I was having the same number linked but it was your bank which asked me to have separate numbers for both accounts.

@sonialok Hi Alok, having the same mobile number registered with both your HDFC Bank account and your Aadhaar (UIDAI) is crucial for security, compliance, and convenience. Recent regulatory changes are designed to prevent bank fraud by ensuring that the person controlling the phone number

People designing such systems must be most velas in their life. Why the hell do I need to have same mobile number connected to bank account as the aadhar. @HDFCBank_Cares @HDFC_Bank

@HDFCBank_Cares That is exactly the reason why many folks don't want the same mobile number linked to bank account vs what is linked in aadhaar. I don't understand how same number improves the security posture when a single mobile hacked is actually easy for attackers to give them access to otps

@sonialok used for banking is the same person linked to the Aadhaar identity. -Ajay, Service Manager

@miten @CRED_club Invite please. Not able to dm.

building Surplus ₹₹₹ to buy the dip! DM for invites.