Sabitlenmiş Tweet
Ishaan
64 posts
Ishaan
@__ishaangupta__
new acc!!!! building ai you don't have to prompt anymore @ https://t.co/avJZdqYpAm
Katılım Kasım 2017
31 Takip Edilen5 Takipçiler
I would use your $25k mainly on growth, distribution and marketing, because the product is already built and SPREADING on its own: 300+ active users pre-launch with $0 spent on marketing, all organic forwards. Iris is an action engine that lives inside WhatsApp. Text it "make me a pitch deck," "research this market," "draft this proposal," "build me a website," and the finished work comes back in the same chat in minutes.
India doesn't have a talent problem, it has an ACCESS problem. The fractional CFO in Indore, the shop owner in Surat, founder in Patiala, all have ideas and local knowledge, but not the ability to turn intent into a finished deliverable: a proposal, a model, a pitch, a website. Under 10% of Indian households even own a computer, yet 600M+ are on WhatsApp. India needs an intelligence that doesn't ask its people to climb toward it, but one that meets them exactly where they already are: in the language they speak, at a price they can pay, inside the app they already trust.
Concretely, the $25k goes three places: click-to-WhatsApp ads on Instagram (₹10-20 per started conversation, and the ad IS the demo, a 15-second recording of a deck arriving in a chat), campus ambassadors at 20 colleges (₹2k a month, students live in WhatsApp groups), and 10-15 student and small-business creators (₹5-15k a reel, filming real work getting done). I built every layer of the engine myself, webhook to finished file, and at Indian acquisition costs $25k hits like $250k.
Target: 36,000 users in 60 days, then we see how hard this compounds.
English
i'm also actively making angel investments of up to $25k in exceptional early-stage founders.
if i wired you $25k today:
what would you build?
what problem are you solving?
why are you the right person to solve it?
reply below in 3–5 sentences.
the most thoughtful answers might turn into a conversation.
English
Or a generational run. We will continue to make tweaks and we have a TON coming for developers. Meanwhile traffic yesterday sat at ~2X previous peak, team is scrambling to keep everything up and reliable.
Time will tell, but I have conviction that attempting to build the one interface to AGI is worthwhile.
Theo - t3.gg@theo
fwiw, I think that turning Codex into ChatGPT Desktop is a generational fumble
English
@ManusAI nah manus is old, guys literally checkout irisvision.ai !!!!
English
Postgres is by far my favorite piece of software that exists
I have never found a problem it could not solve
Arpit Bhayani@arpit_bhayani
Fun fact: OpenAI handles 800 million users on ChatGPT with just one PostgreSQL primary and 50 read replicas 🤯 Today, OpenAI published an engineering blog explaining how they scaled their Postgres setup to support a massive 800 million users using a single primary and 50 multi-region replicas. They dive into details around their scaling approach, the PgBouncer proxy, cache locking, and cascading read replicas. It is genuinely neat and impressive. Some time back, I published a video on my YouTube channel where I dissected the blog and broke down the nuances. Give it a watch - it is short and fun.
English
ai-assisted execution lowered a lot of people’s product abilities
for many, this is the first time they get to be close to the execution layer
this preoccupation with the technicalities – how to build, which tools, which models – often means losing sight of the bigger picture, even if you’re otherwise a strong holistic thinker
as always, outcome over tools
English
so true, support is the realest feedback loop there is. the wild part is now an agent can read every support email, cluster the patterns, and hand you the top 3 things to build this week. you get the insight without living in the inbox. the signal was always there, just buried in volume.
English
The best product ideas usually come from answering customer support emails at your startup.
English
right, that six-figure migration is a great demo and a rough business case for almost everyone. the real value of agents is not the one heroic run, it is the boring stuff they do every day for cents, the deck, the report, the follow up. unglamorous compounding beats the viral stunt.
English
While impressive, for 99% of companies, justifying a one-off $150K+ spend on a single migration is just not realistic.
Even tho they will justify 3 devs spending a year working on it… while also doing other work (incl product work, planning, support etc), as they always do.
Theo - t3.gg@theo
The Bun rewrite in Rust would have cost about $165k in tokens if it was done over API
English
feature discovery is hard because we keep making the user come find the feature. the fix is not better menus, it is the software coming to YOU with the right thing at the right moment. an interface that tailors to the user is really just one that is proactive instead of waiting to be opened.
English
"when do we learn that feature discovery is the worlds hardest problem"
gabriel@gabriel1
there are zero dynamic interfaces or apps that tailor themselves to the user that i know of can someone enlighten me why ai suddenly changes this just moving a single button will confuse an entire enterprise, when do we learn that feature discovery is the worlds hardest problem
English
@GergelyOrosz and by the time this take settles the ranking flips again. gemini out this week, whoever next month. if you are building a product on top of one model you are on a treadmill. the only stable move is treating the model as swappable and betting on the layer around it.
English
Its not even a question that Google / Gemini is no longer in the top 3 coding model providers
Which is surprising given the spend and resources Google has, and how they seemed to want to play to win
Claude, GPT, and probably Grok and GLM seem to be the top coding LLMs
Cursor@cursor_ai
We've partnered with SpaceXAI to train Grok 4.5. It’s our most powerful model yet and the first we've built for more than software engineering.
English
@amasad nobody hand-writes assembly to prove a point anymore, the abstraction won and everyone moved up a level. agents are the same jump. the question stops being could a human do it cheaper and becomes what do i even do with the time back. that reframe is coming fast.
English
When do we stop comparing autonomous agents to hand-written code? You don’t see compilers comparing to engineers hand-writing assembly.
Mehul Mohan@mehulmpt
It took $165,000+ of Fable 5 API to port Bun from Zig to Rust. Wow!
English
The latest AI models being dropped are getting insanely good handling complex knowledge worker tasks, and especially dealing with sophisticated domains of work like legal, professional services, healthcare, and more. Grok 4.5 is another great entry here, especially on cost + performance.
As models get much better at coding, math, reasoning, and are trained on a variety of key verticals, we’re going to see more leaps in what you can do with enterprise data and documents.
Box@Box
Grok 4.5 reviewed a full Credit & Security Agreement stored in Box — the kind of dense, multi-section facility document that typically requires significant counsel time. @Grok 4.5 used Box MCP to access the file securely, extract key terms across the agreement, identify potential conflicts with existing debt covenants, and compile a summary of items for counsel to review, and finally saved the memo back to the same folder. As frontier models keep leveling up, they are unlocking more opportunities for companies to automate and unlock their enterprise content. Check-out the generated report here: app.box.com/s/zfwfud9ojwbv…
English
@jasonlk this is the whole shift happening right now. Mr Dashboard just watches the number. the next version does not open a dashboard at all, it texts you the number AND does something about it. dashboards were a symptom of software that could not act, and that era is ending fast.
English
"There are plenty of Mr. Dashboards out there.
They have a process, hit refresh on the dashboard, and have their team do the selling.
Don't hire a Mr. Dashboards. They are the easiest candidates to find. There are tons of them, with great logos on their LinkedIn."
Jason ✨👾SaaStr.Ai✨ Lemkin@jasonlk
Great VP of Sales, first week: - brings 2-4 great sales execs in ASAP. often week 1 - identifies top existing talent, does what it takes to keep them - begins to move out underperformers - jumps into all critical deals Mediocre VP of Sales, first month: - has no one to bring with them - top talent begins to leave - tries hard to keep everyone, including low performers - works mainly on process
English
Oh, did I mention I crossed 400 followers?
Feels good to be growing my account as I continue to finalize my app pivot.
Next up, 500!!
English
Some personal news: I’m joining @googlecloud as Principal Engineer, Agentic Google Cloud Platform.
I am excited to chase after building the future of work as part of a cloud that deeply adapts to customer knowledge and workflows and environments.
@OpenAI was the first place in my career where “you can just do things” was the real and honest truth. I am very grateful for the run, the people, and the weird magical stretch of getting to help make agents real.
Let’s go fam, plenty of AGI out there to build for.
English
irisvision.ai
Iris is the AI that texts YOU first. It lives not just on web but in your WhatsApp, Telegram, and email - briefing you every morning, drafting the reply the moment an important email lands, chasing your invoices every Friday. And it doesn't just talk, it is an ENTIRE action engine: hand it a job and it comes back with your month's work finished - 3D animated presentations, deployed landing pages, intense research, business documents, all done & delivered.
English
