Ahmer Sultan
4.8K posts

Ahmer Sultan
@cyphorous
I help founders turn ideas into real products. Validate your SaaS faster, and know if an idea is worth building. 18+ years of building products / CTO
Katılım Mayıs 2008
217 Takip Edilen487 Takipçiler
Sabitlenmiş Tweet

@Jason It’s an evolutionary stage. Learn the tools yes. But also develop decision making at scale. As AI doing a lot of work humans still need to review and decide about certain things.
English

For hundreds of years, business owners have replaced humans with machines that don't complain or get tired.
It's now happening at an accelerated rate.
There is only one solution: learn the tools.
If you can't find a job applying to them, start a company.
We're going to see 100 times as many small- to mid-sized companies in the next decade or two. x.com/Overlap_Tech/s…
English

@danshipper @every understood. so I am assuming after review if a fix is needed, either the manager or the technical person themselves can do the fix using AI.
English

@cyphorous @every The technical people on the team use AI to review!
English

We’ve automated every single thing we can @every with AI agents.
And yet there’s way more human work to do than ever. We’ve gone from 4 -> 30 human employees since GPT-3.
I wrote a report on the structural reasons: how AI makes expert competence cheap, why that drives up demand for experts, and why the dynamic only intensifies as we approach AGI.
After Automation: every.to/p/after-automa…

English

@thijsmakes @SimonHoiberg If you can send me a video of that demo, I'll happily take a look on my own time.
Also, What do you mean by AI groups in this context?
English

@cyphorous @SimonHoiberg I'm building something that might fit nicely into that workflow, AI groups + prioritizes customer feedback and recommends what should actually deserved to be implemented. I'm trying to help small teams avoid backlog chaos without bloated PM tools. Want a quick demo?
English

I don't really trust SaaS advice from people who never touch support.
Support is where stuff gets brutally real. Not the roadmap call, not the nice demo, not the friendly customer saying everything is great and then silently churning 3 weeks later.
The most annoying tickets usually point straight at the product work nobody wants to deal with, which is inconvenient but really important.
English

A blackbox is a system (an app, smartphone's OS, search engine etc...) that takes your input, processes them (the middle is key), and gives out an output.
This middle part, is unknown in case of a blackbox. Like someone wouldn't care how Airbnb suggests the best possible stay for them, as long as it satisfies their requirements. What happens in their algo is not known, and that's a blackbox to outsiders.
English

@cyphorous @AdrianDittmann Can you explain to me what a black box is?
If it's an info gathering system, it's that ppl have to not care bc there is nearly no way to stay social/keep friends or even get school reports anymore with a flip phone and no internet.
English

Not really, I have projects in ChatGPT and Claude, I start from there.
I add as my context, updated developed / implemented stories in them.
Then I do sparring. Back and forth. Add my own expertise, and come up with this for my clients to give them the best options, and then take a decision or respond to steer the project.
English

@cyphorous @SimonHoiberg Do you manually prioritize and collect feedback?
English

The first two projects are for my clients. I've built those projects for them.
So they are responsible for their own distributions. As part of what I do, I've given them a recommended plan of distribution as per their market, they are working their own as well.
As for b2c, I am still cooking it up in plans, once I have it launched, my plan is simple for now, I'll talk about this journey on this platform as much as I can.
English

@cyphorous @laurencebuilds Three products across B2B and B2C is a lot to manage from a marketing perspective since they require completely different approaches.
Which one feels closest to a distribution breakthrough right now?
English

@marclou This will probably become the new moat. Quality will be a given, but speed will become the differentiator.
English

I feel Cursor's planning mode is still underrated. People need to pay more attention. When I say people, read Software Engineers here.
Small example, that I love about Cursor's planning mode.
These diagrams. I've built a lot complex than these, and it just makes me understand the design / plan so much better.
Well done @cursor_ai

English

@thdxr Elon, anthropic, and cursor, made their moves. I am guessing they are itching as well.
English

Love this.
I am also building in the founder support space with Yuba.
Yuba is a scalable sounding board for African entrepreneurs that helps founders move from idea to clearer strategy, validation, and execution through a guided step-by-step process. The goal is to make personalized entrepreneurship support more accessible and easier to scale.
Would love to get your feedback on it.
Main platform: YubaNow.com
You can also try the journey simulator here: YubaNow.com/journey-simula…
English

Hey All,
I’m trying to #connect with more people building useful things.
I help founders clarify ideas, build MVPs, and rescue AI-built apps before they waste months.
Especially interested in connecting with people around:
→ Founders / Startups
→ MVP Development
→ AI-Assisted Dev
→ SaaS
→ Indie Hackers
Always happy to meet thoughtful builders. Let's grow together!
English

@cyphorous I spend most of my time building weird AI/dev tools, writing about frontend systems, and experimenting in public.
Would love to connect.
English

apologies to all @postbridge_ customers, i am still not self hosted so this outage currently means post bridge posts are also stuck in queue as the API is hosted on railway and only railway right now
waiting on repair for now, but this tale is getting old sadly
Jake@JustJake
It appears Google Cloud has blocked our account, and so some services are unavailable We’ve escalated this to Google and will keep people posted. Deepest apologies.
English

@AdrianDittmann I've stopped watching keynotes. I mean I can't do anything, I have to wait for the punch lines or important items.
I just ask my agents now to give me the summary, only that's relevant to me.
English

I suspect the next Cursor Composer model will be Grok-based, not Kimi-based like the current one. Composer 3 might even be the Grok 1.5T model @xai is currently training on Colossus 2.
At this point, the xAI acquisition of Cursor also looks more and more like a done deal.
Cursor@cursor_ai
Together with SpaceXAI, we’re training a significantly larger model from scratch, using 10x more total compute. With Colossus 2’s million H100-equivalents and our combined data and training techniques, we expect this to be a major leap in model capability.
English

@ryolu_ In debug mode, can we improve work on instrumentation code it writes, when the local environment for a project is running within Docker?
English

i now use:
Composer 2.5 for planning
Composer 2.5 for building & iterations
Composer 2.5 for debugging
a great all-rounder, especially for UI work – gets you in flow with Design Mode in Cursor
Ryo Lu@ryolu_
i use opus 4.7 for planning composer 2 for building & iterations codex/gpt-5.4 for hard bugs all in @cursor_ai
English

@PolymarketMoney It's either going to be great or very bad.
But so far, both of these companies are doing great.
English








