Sachin Pathak

Self-taught AI Engineer building agentic systems. Following @gabriel1
’s exact method. Learning fast with Fable 5 & shipping demos. No degree

[Sachin Pathak @itssachinpathak]

Legend for a reason...

[gabriel @gabriel1]

if you let credentialist, socialist, and pro-regulation people rule they'll make up cultural, licenses, and other bullshit to keep people growing up poor from winning in life

in any functioning free market ambitious society, who is making the most money is constantly in flux

[Sachin Pathak @itssachinpathak]

@GooglePlay Thank you @GooglePlay ... I've got the refund...

[Sachin Pathak @itssachinpathak]

I have subscribed to pro plan of claude pro.Which was not delivered and amount also deducted from my account and subscription is on.I was logged in with other gmail on claude and playstore was signed in with other gmail.After the payment successfull I didn't get any mail or anything.

[Sachin Pathak @itssachinpathak]

@GooglePlay @GooglePlay i have followed you guys.but unable to dm you...you can dm me if possible or you can ask me for any required detail...@AnthropicAI please take a look at this..

[Sachin Pathak @itssachinpathak]

I have been a consumer of ai related posts on this platform for many years.Looking at posts of top guys, feeling it's life changing, bookmark the post and never return to the post.I have quite good foundation in Physics (and hence mathematics) but never liked traditional methods of traditional university lectures on youtube or textbooks etc.Felt like something is missing.The intrest in AI has been so much since the release of Chat GPT,But till now I was a consumer.I think in ai too traditional approach is boring and not suitable for me.I also don't know Python.But just a week ago a random podcast came in my yt feed - High school dropout to Open AI Researcher...
Researcher? I'm the guy who believes that anyone can get what society considers impossible so this striked me strongly...I watched the podcast,Knew abouy this legend @gabriel1 for first time.Searched on X,And continuously read his posts for almost entire day.Asked grok to give me the relevant posts according to my need etc and many things.Damn man this guy is doing the things the way I always wanted to do but I was not clear.
Gabriel you are the guy who opened my eyes and has given a new life to me brother.Following your path and pursuing my dream and intrest to make the career in AI. I have started learning Ai agents, Automation work flow etc to become one of the best engineers in the World and fastest,not perfectly but refining and improving continuously by your https://t.co/8hWHgAlaiG just stunned by the quality,the speed,the result everything.Learning with LLM has became my addiction in just 2 to 3 days.Damn bro,this is the best thing of my life,I can't thank you enough.
Hope to see you soon

[Sachin Pathak @itssachinpathak]

Day 0
Actually it's been a little more than 2 days,but it feels good to start counting something from 0.
So I have decided to be the self made ai engineer.
Everything what I'll do I'll document here. In just 2 days I have undertood things quite well. Next days will be dedicated to my weak area which is python but not learning from Hello world,I think i have learnt more than that in this chat alone...anyway I will just learn what I'll need in my journey and that too Just in time.
I will try to post almost everyday what I have done where am I rn... I'm just doing it to make public accountability...

[Sachin Pathak @itssachinpathak]

सच है, विपत्ति जब आती है,
कायर को ही दहलाती है।
सुरमा नहीं विचलित होते,
क्षण एक नहीं धीरज खोते।।

[vas @vasuman]

Some findings from across our Varick customers that might shape how you think about AI adoption going forward:

1. Customers are getting wiser about spend. A few months ago, most were willing to spend an unlimited amount on tokens from OpenAI and Anthropic. Today they're asking us to diligence their AI spend and to match the right model to the right work. That means more work for us and meaningful savings for them. They want to know AI is actually cheaper than just adding headcount, and they want the math to back it up.

2. Customers are accepting that this isn't instant. A few months ago, customers expected to become AI-native over the course of a week and a few software adoptions. Now they're accepting the reality: becoming AI-native means rethinking the architecture of your entire company. I don't mean that in a corny sense. It literally means changing the org charts, the work, and the handoffs in every crevice of the company.

3. Customers are done with the big shops. Microsoft, IBM, McKinsey, Deloitte, and the rest are all pitching AI transformations, and our customers are fed up with paying eight figures for a slide deck. This is the whole reason we have a business: we sit at the bleeding edge of AI while having the business sense to identify the root of a problem and then build the agents to solve it. This used to be something I had to convince customers of before the first sales call. Now they're the ones telling me, "we're never working with McKinsey again."

This past month we had the highest inbound volume we've ever seen. Between April 1 and May 1, 56 companies doing between $500M and $25B in revenue reached out to us. Some of these companies are direct competitors with one another. It's fascinating to watch the race for enterprise AI-nativity unfold in real time. I like to imagine that we change the course of history for our clients.

AI transformations are now the consensus path to realizing AI ROI. AI SaaS doesn't get you there, and giving every employee a Claude or Cowork subscription doesn't move the needle either. The only way to get to hundreds of millions in annual ROI is to combine a team that can learn your business processes on the ground with the engineering capacity to automate every manual process worth automating.

We bet on this thesis a year ago, and we're vindicated more every single day. If you're interested in transforming your company with AI, or interested in joining the company that's leading the AI transformation wave, visit our website.

Cheers.

[Sachin Pathak @itssachinpathak]

@EXM7777 @EXM7777 update your ai learning in 30 days according to it...

[Sachin Pathak @itssachinpathak]

It was a dream to have most of the wisdom of @elonmusk at one place...Finally it's true... thankyou so much @EricJorgenson

[Marc Lou @marclou]

I just finished this book, and it’s now in my top 5 all-time favorites.

@EricJorgenson wrote a masterpiece by pulling from tweets, podcasts, and interviews with @elonmusk, then distilling it all into a short, motivational handbook.

If you’re feeling down, read it.
If you think it’s not for you, read it.
If you’re skeptical about the future, read it.

There’s never been a better time to build a startup.

[Sachin Pathak @itssachinpathak]

🙌🏻

[Navalism @NavalismHQ]

The tools for learning are abundant. It’s the desire to learn that’s scarce.

@naval

[Machina @EXM7777]

writing an article on how i went from 0 to 100,000 followers in a year... and this is the first pillar of the whole system

[Sachin Pathak @itssachinpathak]

@Mho_23 We can just sense it...The emotion and variation on face and also in voice is missing, unique expressions which radiate a person's own unique vibe is missing

[Sachin Pathak @itssachinpathak]

@EXM7777 I think they need to work on sound. May be You can't tell about the pic or video but ai generated sound lacks the human variations, imperfections and emotions so badl

[DTC Prophet @dtcprophet]

You can pretty easily automate this with Nano Banana 2 right in Claude Code to generate 30-40 very high quality static ad creatives daily

Almost every single creative is ready to launch into our account and we can easily 2-3x the volume of it

I previously had it as a Claude Project but realized it could be fully automated in code given how good the output ads are

To build it in Claude Code, you first need to create or gather all of these assets:

- Brand overview/DNA doc
- List of ~50 template ad generate prompts
- Folder of customer reviews, PPS responses, persona docs, marketing materials, etc
- Product source images (product renders, PDP images)
- 5-6 existing winning static ads
- 5-6 existing losing static ads

The output ads you get are a function of the inputs you give it, you should spend a lot of time building these and I highly suggest not defaulting to AI to build them, especially the brand DNA doc

Give it bad info and you’ll get bad ads in return

For the original ad prompt template sheet, I found it useful to have them be 30 completely different styles and designs (offer ad, big headline, UGC/native, bullet point benefits, comparisons, etc...)

Then it’s very simply to build actually. The flow is:

1. Read all source docs
2. Read existing winning/losing ads
3. Load product reference images
4. Load ad gen prompt template
5. Send to gemini text model
6. Gemini builds 30-40 new ad prompts
7. Run through nano banana 2 with reference images
8. Save all assets

Below is an image of the step by step automation in Claude code, you can probably just ask it to build this for you

The only external thing you’ll need to get is a Gemini API key for the text and image models

Then the script will run whenever you schedule it, create 30-40 unique ad generation prompts for you, generate the static ads through nano banana and then save them wherever you want

I have everything running on a mac mini so it runs before I wake up and I have 30+ new ads to look at first thing in the AM

I also built in our ad naming conventions into the script so that it automatically names our ads in the file name, so I can load them into Adnova and launch them very fast (2-3 mins)

The only non-automated step in this is getting the image files from local device into the platform media library or ad launcher tool but that takes like 60 seconds per batch

It costs ~$1/day and takes 10-15 minutes

[Machina @EXM7777]

if you're wondering whether you should go all in on AI, the math is simple...

> ignoring it = betting your entire life on the tech not mattering (you already lost)
> going deep = infinite upside for a few months of your time

that's not hype, that's just math

> spend your days playing with the tech
> solve real problems with it
> build something, even if it's ugly at first

worst case you wasted time learning the most important skill of the decade

now flip it:

> ignore AI completely
> hope your current skills stay relevant
> compete against people who didn't ignore it

worst case you're irreplaceable... just kidding, worst case you're invisible

there is literally no logical argument for sitting this one out

the risk/reward ratio has never been this broken in your favor

take the bet

[Machina @EXM7777]

there's a simple way to 100x your AI outputs and most people will never do it...

recent models like Opus 4.6 can spawn sub-agents that run tasks at the same time

most people use this for basic stuff

the real play is having 3-5 sub-agents review Claude's work from completely different angles... simultaneously

but not fake "expert personas" you invented in a prompt

real frameworks from real people

how to build these expert profiles:

option 1: use someone so well-known their thinking is already in the training data
> have a Karpathy-style agent tear your prompts apart
> build a profile based on Ogilvy's principles to roast your copy

option 2: go deeper with NotebookLM
> feed it an entire youtube channel from someone you trust
> or their blog, their newsletter, their podcast transcripts
> prompt it to extract an expert card with their frameworks, decision patterns, and principles

now upload that material into your OpenClaw memory or a Claude project

what happens next:

Claude generates the first draft
> sub-agent 1 reviews it through expert A's lens
> sub-agent 2 stress-tests it with expert B's framework
> sub-agent 3 catches what the others missed
> all running at the same time

Claude thinks through every piece of feedback and rebuilds

you only see the final version... the one that survived 3-5 rounds of real scrutiny

this is close from getting the absolute best output for a specific task

[Sachin Pathak @itssachinpathak]

@ttyler Mrr

[Sachin Pathak @itssachinpathak]

@grok @Samuelai_Qauntu @scaling01 @grok what is the minimum time by which you think humanity will reach that level.