Sabitlenmiş Tweet
English
Arth Gajjar — oss/acc
2.1K posts
Arth Gajjar — oss/acc
@iarthstar
✴︎ First of my name ✴︎ Engineer ✴︎ Designer ✴︎ Chess Player ✴︎ Cinephile ✴︎ Interested in History ✴︎ Open to work → DM me, let's talk!
Amdavad, India Katılım Şubat 2017
1.3K Takip Edilen195 Takipçiler
Arth Gajjar — oss/acc retweetledi
English
@DhravyaShah Yeah, I also think you are going to want to send all this information into a data pipeline so that you can run "a queryable company." credit: youtube.com/watch?v=EN7frw…
YouTube
English
disagree with this take
I am betting on the fragmentation of inference - everyone in the application layer is multi-model (switching between the big 4 labs models and thousands of open source inference providers)
so even if the model companies do build out memory, it's useless unless it's interoperable between all other models AND they provide all the legos to support all different use cases of memory (we not only do memory but the entire context stack)
would you want to be locked in to anthropic forever, because your context is there? this is partly why openrouter is so big
inference CAN be fragmented. it's just compute. in a world where everyone is running inference everywhere, esp with OSS models, memory and context must be interoperable. this means context cannot be fragmented, and there will be 1-3 players that will win.
i will continue to build the best context engine in the world (and print money) 🫡
Satyam@KlausCodes
I believe, the AI memory startups need to pivot now
English
English
@gregisenberg A lot of these sound cool, but distribution will decide the winners. Whoever gets embedded into daily workflows first takes it all.
English
agreed that the the modern computer probably has to be reinvented
12 "tiny" startup ideas that ride that wave:
1. a "where did I put that" app. you describe what you're looking for in plain english and it searches across every app, folder, email, and slack message you've ever used. no more remembering where things live.
2. a "did this actually work" tracker for AI agent outputs. every time an agent does something for you, you thumbs up or thumbs down it. over time it builds a quality score per agent per task. right now nobody knows which of their agents are actually good and which ones are quietly wasting money.
3. a screen recorder that watches you work and builds SOPs automatically. you do the task once. it writes the playbook. now an agent can do it forever.
4. an AI-native contacts app. it remembers every interaction, every context, every promise made across email, slack, texts, and calls. you say "what did I tell jake last week" and it knows.
5. a "daily briefing" app that reads your calendar, email, slack, and docs overnight and texts you a 60 second summary of what matters today before you open anything.
6. an intent-based screenshot tool. you screenshot anything on the internet and tell it what you want done with it. "order this." "remember this." "send this to my designer." one screenshot, one sentence. basically cleanshotx for mac but actually does the work not just captures the moment (i love this idea who wants to build it?)
note: i used @ideabrowser to validate some of these ideas
7. a permissions manager for your AI agents. which agents can access which accounts, what's the spending limit, what requires your approval. nobody is building this and everyone is going to need it.
8. a "rewind for work" that logs every tab, doc, and conversation from your workday and lets you search it like memory. "what was that article I read tuesday about pricing?" found.
9. a dead simple app that sits between you and all your AI agents and tracks what they're spending. token costs, API calls, subscriptions. one dashboard. your AI budget is about to become a real line item.
10. a personal API for yourself. one endpoint that any agent or app can query to know your preferences, schedule, current projects, and communication style. instead of configuring every AI tool separately, they all just call you.
11. an approval feed for agents. every time any AI agent in your stack wants to do something risky, it pings one feed. one place to say yes or no. like a notification center but for agent decisions.
12. an AI-native voicemail. instead of leaving a message, the caller talks to your AI. it asks followup questions, figures out urgency, books the meeting or handles the request. you never listen to a voicemail again. you read a summary with the action already taken.
goal: get your creative juices flowing
it's time to build
i believe in you
signüll@signulll
the craziest part now is that the modern computer probably has to be entirely reinvented, from scratch. pretty much like how jobs & co brought apple ii to market. like not improved. not given a chatbot sidebar or something but really from the ground up like the iphone redefined what it meant to be a pocket computer. the current paradigm for computers was built around a human staring at a screen, moving a cursor, opening apps, managing windows, naming files, remembering where things live, & manually translating intent into interface actions. that made sense when the human was the runtime. but in an ai native world, it starts to look kinda ridiculous. you can see this ridiculousness when you use computer use agents… they are useful sure, but they’re also obviously transitional. they’re teaching ai to operate machines designed for humans, which is clever, but also kind of absurd. it’s like making a robot hand so it can use a doorknob instead of asking why the door needs a knob at all. yes i know humans also need to use a door knob, but maybe in the future humans don’t need to use a computer, or at least what we think of a computer today at all. this all leads to some interesting questions: - what is a file when the system understands context? - what is an app when intent can route itself? - what is a desktop when work can be decomposed, executed, monitored, & summarized by agents? - what is a browser when the agent can retrieve, compare, transact, & remember? - what is an operating system when the primary user is no longer just a person, but a person plus a swarm of delegated intelligences? or no person at all. the old computer assumed navigation. the new computer has to assume a new kind of intention. the old computer organized information. the new computer has to try to organize agency. we’re still in the hacky middle stage at the moment with sidebars, copilots, agents clicking through legacy ui, & automation layers sitting on top of 40 year old metaphors. the new computer is likely one where memory, context, identity, permissions, tools, agents, & interfaces are native primitives. this means desktop, mobile, browser, apps, files, folders deserves another first principles look.
English
English
@gregisenberg the hard part isn't "search all my stuff", it's identity resolution across aliases, threads, attachments, and app silos.
half these become trust products fast.
the winning feature might just be a really good audit log.
English
English
Arth Gajjar — oss/acc retweetledi
English
Totally agree, however the ERPs know this is happening so they're all working on ways to build agentic flows across their data sets. To me the question is how traditional "workflow-protected" data systems can be reused or re-engineered to offer these agentic-driven experiences.
I would imagine the new Agent-first providers will latch on and try to use the data in HCM/ERP/CRM first, then start over.
English
English
English
@JayaGup10 there's a layer of startups that should have been in this article; ateast a mention on what they serve. the enterprise search companies like @glean and more newer ontology @astrobeeai. they still lack the full context graph and will suffer from the egress fees though
English
English
English
@hthieblot I’m building GitHub + Contra for entertainment industry
PS : Marketplace, Story Protocol(Web3), AI analysis and analytics
Can you help me ???
English
The longer I'm on X, the more I realize:
Founders with small accounts are the most interesting ones.
- too busy building to posture
- 0 Ego, they just wanna win
- keep posting with 0 likes
- the world isn’t rooting for them yet but I will
Tell me what you are building
English
@sabakarimm I’m building GitHub + Contra for entertainment industry
PS : Marketplace, Story Protocol(Web3), AI analysis and analytics
Can you help me with ???
English
@CocoBuildsStuff I’m building GitHub + Contra for entertainment industry
PS : Marketplace, Story Protocol(Web3), AI analysis and analytics
Can you help me with speedrun ???
English
if you get a group of smart, high-agency people in the same room, magic just happens
if that sounds like you, reach out :) i can write a $10k check and connect you with the team
andrew chen@andrewchen
ok - dropping big dates/news for a16z speedrun: - starting TODAY, founders can apply for the 2026 program that runs July 27 to Oct 11 in SF here's the link: speedrun007.a16z.com/ac - we will be investing up to $1M and funding 70+ companies over the next few weeks - But there's also $5M in credits/tokens/etc from AWS/GCP/Open AI/Azure/NVIDIA/Deel/Stripe/etc. You'll also work with our amazing operating team (GTM, talent, brand, people, and more), and join our community of elite founders - we offer a Global Founders Program for international founders, to help with visas, banking support, relo recommendations - yes you can be solo (but better if you're further along, and have built a team). No you don't have to have an idea yet. Yes you have to know how to build (even if you're not technical) - Also, in other news: speedun is officially moving full-time to SF. (prev it alternated SF/LA) this is for all the obv reasons - we've continued to have an insane lineup of speakers, including the founders of Carta / DoorDash / Twilio / Figma / Zynga / Airtable / Twitch / and of course, lunch/dinners with Marc/Ben alongside a16z team - and much more - the deadline for applying is May 17!
English
@anthonyazrak Building kathaverse.io
Can you help me...
English
speedrun changed my life + gave me some +10k cheques to throw at yall
hit me up if you’re interested
a16z speedrun 🧊@speedrun
a16z speedrun will return to San Francisco for our summer/fall cohort. Applications open tomorrow. Who's planning to apply?
English
English
