Reality of building in public :
you can build anything now, have a public audience as well.
that doesnโt mean anyone will want it.
English
๐๐๐๐๐๐ ๐๐๐๐๐ // ๐๐๐๐
365 posts
๐๐๐๐๐๐ ๐๐๐๐๐ // ๐๐๐๐
@sierra_tech_
spent 4 years in product design. now just chilling
LA, CALI Katฤฑlฤฑm Aฤustos 2025
131 Takip Edilen56 Takipรงiler
@Prathkum the task got outsourced, the judgment didn't, that distinction is going to separate a lot of careers over the next few years
English
Your first 100 customers should know your name.
Scale comes after intimacy, not before.
English
@sierra_tech_ @xwanyex And it's more like 2x-3x increase if you're talking production apps at scale, where just "writing the code" isn't the limiting factor.
English
@stats_feed spending millions in compute because people are being polite to a chatbot is genuinely one of the more unexpected infrastructure problems in tech history
English
Sam Altman admits that saying "Please" and "Thank You" to ChatGPT is wasting millions of dollars in computing power.
English
@Prathkum cursor didn't kill vs code; it just inherited the body and gave it a much higher IQ. The transition from "text editor" to agent harness happened so fast we barely noticed the switch
English
@AndrewYang the productivity gains have to go somewhere, the question is whether the political will to redistribute them shows up before the displacement does
English
Itโs clear that AI will wind up funding universal income. Letโs make that happen ASAP.
English
@Prathkum because we have a model launch every other week
English
@techxutkarsh an AI that can actually use your computer instead of just chatting with you is the product everyone described wanting for years and nobody could ship until now
English
Perplexity just launched something new called Personal Computer.
In simple words: It can read your files, use your Mac apps, and help you do things on your computer.
It keeps running all the time on a Mac mini, so itโs always ready.
And you can control it from your iPhone anytime.
Basically, itโs like having an AI that can actually use your computer, not just chat with you.
Perplexity@perplexity_ai
Today we're releasing Personal Computer. Personal Computer integrates with the Perplexity Mac App for secure orchestration across your local files, native apps, and browser. Weโre rolling this out to all Perplexity Max subscribers and everyone on the waitlist starting today.
English
@dougboneparth the pivot to AI compute was the right call, margin profile on lemonade was never going to support a Series A anyway
English
My daughterโs lemonade stand was struggling, so we pivoted to AI compute.
Sheโs been acquired by Microsoft for $1.2 billion.
English
Hard to build trust in the technology when the people building it don't follow their own rules
English
@adityaag Moving the goalposts is a classic coping mechanism for tech shifts. At this point, denying the utility is just a commitment to being less productive than the rest of the world
English
The arc of a LLM skeptic:
1. They are just auto-correct
2. They are just next-token predictors
3. They hallucinate
4. They can only be useful for the data they are trained on
5. They are only good for coding
6. They are only good for verifiable things
Just admit it! LLMs are awesome.
English
@trikcode the one platform that desperately needs disruption and somehow survives everything, atp LinkedIn is protected by the fact that everyone hates it equally
English
๐๐๐๐๐๐ ๐๐๐๐๐ // ๐๐๐๐ retweetledi
โMade without AIโ is the new prestige label.
English
@sytaylor The DraftKings alert is a "brutally honest" feature moving from static tracking to automated, reactive workflows based on real-time transactions is exactly what personal finance has been missing for a decade
English
Perplexity and Plaid built your personal CFO.
Beta testers are building:
- Student loan repayment plans.
- Subscription trackers that cancel unused service
- Getting help filling their taxes with almost zero effort
One guy told Computer to email his wife every time a DraftKings transaction hit his credit card.
English
@testingcatalog so the pivot to computer wasn't a bet it was already working
English
Perplexity grew from $100M to $500M with only 34% growth in team size and fully pivoted to Perplexity Computer.
Computer for businesses ๐
Aravind Srinivas@AravSrinivas
Perplexity started as a small business tool for ourselves. We had 4 people and no revenue with AI at our fingertips. The pivot to Computer is actually a full circle. Founders are using it to grow companies that matter to the economy and their communities.ย Itโs rewarding to see it now powering small businesses and startups in big ways. Perplexity is still a startup. We just 5Xโed revenue from $100M to $500M with only 34% growth in team size. 2x revenue growth in 2026 with same small team. And weโre just warming up. Everyone here works at a small business, and everything we build is for people who build.
English
@iam_chonchol @perplexity_ai dashboards assume you know what question to ask before you open the app, the prompt interface assumes you don't and meets you there, that's a fundamentally more honest starting point for most people
English
Perplexity isnโt building a finance app.
Theyโre removing the need for one.
Every fintech tool you listed is built around interfaces โ dashboards, charts, categories, workflows.
@perplexity_ai flips it: the interface is the question.
That changes everything:
No onboarding friction
No learning curve
No rigid structure
Just raw financial context โ instant answers.
And once all accounts are connected, the moat isnโt UI anymoreโฆ itโs data + intelligence layer.
If they nail trust + accuracy, this becomes the default financial OS.
Dashboards were phase one. Prompts are phase two.
English
@paraschopra The verifier problem is the real issue for most labs. Mastering low-causal domains first makes sense, but high-complexity tasks are the true end game
English
The most useful way to think about AI capabilities is to think of target domains in terms of 3 orthogonal axis:
- Ease of building a verifier (coding is easy, lab equipment manipulation is hard)
- Causal complexity in terms of number of confounds (math problems is low as answers donโt depend on external factors, startup success is high complexity as it depends on many random factors)
- Economic attractiveness (high for coding, low for many domains)
What weโve seen the first to be automated are domains where building verifiers is easy, causal complexity is less and economic attractiveness is high.
No wonder coding is the first one to see big jump.
But inflated valuations of AI companies need to be justified, so I fully expect that these companies will keep attacking the next best domain they can until they exhaust the economic attractiveness constraint.
English
๐๐๐๐๐๐ ๐๐๐๐๐ // ๐๐๐๐ retweetledi
