Built From Blind

Claude rated my own progress after 4 days. Systems thinking: 5 ⭐ Execution consistency: 4 ⭐ The system is built. The log exists. The diary exists. The only variable left is whether I fill Tab 3 on a tired Tuesday after a bad session.

@Ishant23April My V0 mistake was using a random 'token saving instruction'. Glad I noticed it did block the outcome. Solved in V1. Will check on details your post if I can improve a bit more without sacrifying quality 🙏

reduce token usage

@TheOneGroupAI This botleneck make sense. For now I control every step of my PM as every step matters. I still win so much time.

Qodo raised $70M because AI writes code faster than humans can verify it. The SMB lesson? Automation creates a NEW bottleneck: quality control. Whether it's AI-generated content, code, or customer emails—you still need a human in the loop. Build your process like this: AI generates → Human verifies → System deploys Speed without accuracy is just faster mistakes.

@2NLChillGrind Top 10 out of 3k, thats a real great result even if frustration always happen until the win. Congrats 🎉

Maybe I'll stick to online... Final table, denied a double up and a bounty with my AKs getting beat by AJo. Then with 10bb left on the button. CO min raises (has me covered). I jam all in with A7s. Villain calls with T8o. They hit 2-pair on the flop. Savage last few hands but a good run nonetheless.

@NaivaidyaY66600 I'd say staying consistent. Exciting to start, then somehow difficult to do the things again and again when needed.

What’s the hardest phase of building? - Starting - Staying consistent - Getting users -'Making money

@raahilgadhoke Feel that it happened to me this past weekend too, and foe the next X months ^^ GL

I went from not knowing what I cared about to not having enough time to do all the experiments I care about now. Good problem to have.

I never considered myself a project manager, more a team one. But I discovered Claude. And I liked it. Instead of asking to do things, I am building a project manager, my first real AI project. I have a clearer vision for the next 12 months than I've had in years.

@IanLandsman Are you IN or you just think its gonna happen? Could come from one firm that already own them, with tools a d budget, or from a random builder. Your take?

Amazingly 20yrs in there's still room for a social network laser focused on creators, collaboration, and information dissemination. Someone is going to crack the code eventually.

@vivekps143 I am a newbie, looks interesting. Would you explain it in a few different words so I can reply to 'would you use it?'

@AnuKrishna_ @framer Love challenges. What do you need? A Jury will rate the best delivery with this tool? A few words to explain the challenge?

just got approved for #FramerChallenge Season 2 tried last season. didn't hit the milestone. this time i'm finishing what i started. @framer $10K by July 5th. let's go.

@VanjaPoker Mental has been my strenght back in time, though mostly in micros, in a different life stage. I restarted recently and I am taking care of 'how to' now. While big € swings might be harder now (lost training), my expectations of poker are different now, thats my mental anchor.

In 2025 i have been on the worst downswings of my life. Lost sleep. Questioned everything. Here’s what I learned about the mental side of poker: 1. Variance is not your enemy, your reaction to it is. Bad runs are part of the game. How you respond to them determines everything. 2. Stop results-oriented thinking before it starts. One hand tells you nothing. One session tells you nothing. Judge your decisions, not your outcomes. 3. Tilt has layers. There’s obvious tilt. Then there’s subtle tilt, playing slightly too loose, slightly too passive. Learn to recognize both. 4. Your emotional state is a leak. If you can’t identify when you’re off your A game, you can’t fix it. Start tracking your mental state, not just your results. 5. Losing sessions are information. The best players use bad runs as diagnostic tools. What did this run expose in your game? Swings often exposé big leaks. 6. Detachment is a skill, not a personality trait. You can train yourself to care less about individual outcomes. It takes reps, just like any other skill. Accept what happened, allow yourself to feel the emotion and let go. 7. Taking brakes can be helpful Taking your mind of things by going out with friends, doing whatever you like to do helps to get you back on track. 8. Build a pre-session routine. Five minutes of intentional preparation beats hours of autopilot play. Get your head right before you sit down. 9. Study your worst moments. Not just bad beats, moments where you tilted, spewed, or quit too early. Your mental game leaks cost you more than you think. 10. Poker is a long game. Play it like one. Your results this week are irrelevant. Zoom out. Bonus: Doubling down on study and having a proper BRM helps a lot with confidence. The mental game is the most important skill to have, invest in getting better You got this champ ❤️

@ProcIQAI Interesting. I build for myself only, but will add to handoff the security part, mentioning memory. Got my major update ongoing, from V0 to V1, implemented deeper memory, thanks to your previous post 🙏

This is the research everyone building agent memory needs to read. 90%+ success rate hijacking agents via poisoned memory entries. The takeaway: if your agent learns from outcomes, those outcomes are an attack vector. Memory systems need the same security rigor we apply to databases. Not the afterthought treatment they currently get. Top of mind for us at MemLayer. x.com/NYsquaredAI/st…

@BlackRainPoker Start tight aggressive. Cant learn to play bad hands before learning to play good hands.

Beginner mindset shift: Every fold is a win. You didn’t lose a hand. You saved chips.

@ProcIQAI I will share progress here, a mix of lessons learn and real 'tech' steps done. I'll be happy to read more from your projects too.

The token limit is exactly why a separate memory layer matters. Trying to fit all past context into the prompt breaks fast. Better approach: log outcomes to an external system, then retrieve only what's relevant for the current task. Persistent learning without burning your context window. That's the core idea behind MemLayer. Would love to hear how your PM project evolves.

Most agent frameworks focus on orchestration. What to run, when, in what order. Almost none focus on what the agent learned last time it ran. You wouldn't hire a contractor who forgets every project they've ever done. But that's exactly what we accept from AI agents today. Memory isn't a nice-to-have. It's the difference between a tool you babysit and infrastructure that compounds.

@ProcIQAI For now I use snapshots after each section, make a brief possibility, but this brief part is different that proper exportable logs. I just finished my first testing, and I am one step away from my major post testing update, need to make sure I have good quality logs.

@draft_ofkritika Thats why I am having fun with AI with my unused algorythmic classes 20 years ago. I always knew it would be useful.

do you agree?

@IanLandsman There are always lessons to be learned. They did wrong building this way overconfident. What did we not anticipate? 🤔

Hot take. Claude is doing the right thing. Give the real customers a better experience (keep infra working well for them) while ditching people having their bots loop over their calendar 100,000 times a day.

@sarwangsinha The AI fall, not for AI, but for the one's that did it wrong. You mean even after? 😅

Everyone keeps asking what comes after the AI boom. In my view, its robotics.

@SaiAvala Looks like a good winning session. Well done.

Amateur poker status April 4, 2026: 40.3bb/100 single table NL10