1. X
  2. Anish Acharya
Log inSign up
Anish Acharya
a16z
3,872 posts
Anish Acharya profile banner
@illscience

Anish Acharya

a16z
@illscience
🦞AI Apps investing @ A16Z; A1111; Boards of Ethos, Krea, Deel, Clutch, Untitled, Happy Robot + more; If you’re not at the table, you’re on the menu
San Francisco
open.substack.com/pub/illscience
Joined March 2009
2,867
Following
33.2K
Followers
RepliesRepliesRepostsRepostsMediaMediaArticlesArticles

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • Pinned
    @illscience
    Anish Acharya
    a16z
    @illscience
    Jul 13
    Article cover image
    Article
    The Most Human Technology Ever Made
    Over the weekend, Mira Murati wrote a phenomenal essay that you should read in full: “The future worth building is human.” I think it’s one of the best essays of the year, because it gets at the...
  • @illscience
    Anish Acharya
    a16z
    @illscience
    Aug 29
    0→1 is hard. People will open source you. Thought bois will elegantly pick apart your product choices. Poasters will say nobody will ever use it, but also that it’s a commodity. That you’re the symbol of the top. That you’re cooked cooked cooked. Until you’re inevitable. Then
  • @illscience
    Anish Acharya
    a16z
    @illscience
    Aug 27
    Super thoughtful notes on the “dark arts” of pricing from Sarah & Tugce: - When an application prices work in tokens, it imports the model provider’s cost structure and anchors the product’s value to a unit whose cost keeps falling. - A token-based price trains customers to
    @sarahdingwang
    Sarah Wang
    a16z
    @sarahdingwang
    Aug 27
    Article cover image
    Article
    You are not a model. Don’t price per token.
    Token pricing began in the right place: the model layer. When OpenAI launched its API in 2020, charging for the computation a model consumed was a sensible way to meter raw inference. However,...
  • @illscience
    Anish Acharya
    a16z
    @illscience
    Aug 26
    This is one of the reasons that computer use works so much better when sitting behind an async channel like iMessage .. you want it to be able to work at varying time/ambition scales, just like an employee, assistant or friend .. and obviously this is the worst it will ever be
    @kgauthier15
    Kevin Gauthier
    @kgauthier15
    Aug 26
    Replying to @illscience
    Computer use is great but still expensive and slow. It’s a nice intermediary solution when API/MCP are not available
  • @illscience
    Anish Acharya
    a16z
    @illscience
    Aug 26
    Amazing how quickly the Overton window shifted on aggressive computer use and storing credentials. Payments will be next.
    @nickbaumann_
    Nick
    @nickbaumann_
    Aug 25
    In case you haven't been paying attention, ChatGPT Work (web) now has the following primitives: - its own browser w/ best-in-class persistent auth - a 16gb RAM 9 core computer - plugins to all the apps you use - 5.6 w/ Codex compaction 🤲 here is your always-on personal agent
Advertisement
Advertisement