Pinned Tweet
On my way to learn AI by building an app:
✅ Figure out user flow
✅ GraphQL resolver to upload the CSV
Next up:
⏳ UI (although I like Altair 🤪)
⏳ Job Queue for the AI stuff
I recorded a screencast this time
#code #buildinpublic
teapaw.org/swashata/status/194318…
I have the basic app (Nextjs & my old own GraphQL server) up & ready and have figured out to have AI analyze the statements. Next 👉
- Create API & UI
- Ready the MVP
- 🚀 😅 🙀
I thought I'd share some code sneak peek too 😬
teapaw.org/swashata/status/194103…
Hot take: What if, instead of AI replacing programmers, programmers use AI to replace every other professional? 😁
Renovating a house? Why hire an architect? Just ask Astra.
Hmm, perhaps it really is over. Today for the first time I opened ChatGPT before I opened my terminal or slack 🙀
Nice try @urbancompany_UC , you are now blocked 🤣 Send me a letter when you actually become a phone app to count my missed calls
It's all good and everything. But why create the cross platform? Anyone from those platforms can also make similar wish 🤷
Just an evening rant: If my $100/mo agent can't understand your $10/mo SaaS, then no, I won't use it 😅
Swas retweeted
How I do code reviews these days: more code = more selective human review i.e. don't read all the code.
1. Every PR gets a multi-agent first pass. It should find bugs, verify them, rank by severity, suggest fixes. Approval stays a human call on anything that matters.
2. Low blast radius changes on less sensitive code (there's often a lot!) can skip a deep human review once that review is clean. This helps keep the explosion of PRs manageable.
3. Core / sensitive paths still need an owner and human sign-off. That's where you spend time: verification, constraints and earning trust in what the agents can safely cover. You want to keep recoverability.
Agents do the first pass and humans cover blast radius.
Great question from @GergelyOrosz!
Saw someone say in LinkedIn promoting a strong engineer to lead was a mistake because the team stopped asking questions, the lead solved everything.
What a stupid take!!!!
Coach them to guide instead of solve. Just groom them to become the technical leader.
Oh come on! I like AI a lot, but this is just overstatement on steroids
With Codex, @asana finished a frontend test migration from Enzyme to React Testing Library in two calendar weeks—a project expected to take five more years.
openai.com/index/asana/
ChatGPT is like the fun distant cousin who's very fond of you and always smiling and want to talk.
Claude on the other hand, my oh my 😅
