Manoj Subramanyam

Entrepreneurship is value creation at scale. Building @SwadesicApp

[Manoj Subramanyam @unscriptedmanoj]

I thought laya is India's answer to Jev but it turned out the laya came out first conceptually.

March 2025: published the earlier research
September 2025: a second paper
September 18, 2026: Laya released

Good read
https://t.co/pN4fWjxmFD

[Manoj Subramanyam @unscriptedmanoj]

I was waiting for India to step into the frontier intelligence for a long time. Laya is surely the answer and it is opensource.

Proud of you @Nandakishorm1 for releasing papers a year back and made this model available.

https://t.co/hWrdqe1Y1i

[Manoj Subramanyam @unscriptedmanoj]

@witcheer @HermesWatcher This post only have 8 comments. Missing Mobile app is real deal

[Manoj Subramanyam @unscriptedmanoj]

@galati_janet @ManusAI @Red_Xiao_ @hidecloud @peakji That's bad

[Manoj Subramanyam @unscriptedmanoj]

@opencode Opencode is by far a net positive AI company

[Manoj Subramanyam @unscriptedmanoj]

@Orbily_App @opencode subagents are token hungry.

[Manoj Subramanyam @unscriptedmanoj]

Entrepreneurship is living in a jungle. Lots of options to hunt and to be hunted. Often, we find ourselves swinging between the two.

Love the comfort and safety a zoo can offer. But that's not the way we meant to live our lives. You can get peace of mind with strength too.

[Manoj Subramanyam @unscriptedmanoj]

@heystefan_ That's fun.

[Manoj Subramanyam @unscriptedmanoj]

@0x0SojalSec Interesting.

[Manoj Subramanyam @unscriptedmanoj]

@g9robot @0x0SojalSec @grok https://t.co/jFJJde4hU7 HF

[Manoj Subramanyam @unscriptedmanoj]

@moritzkremb It can't output text right, how did it type Wikipedia?

[Manoj Subramanyam @unscriptedmanoj]

The possibilities just got expanded. What we are looking at is definitely another ChatGPT moment and it is available for us. What do we build is the next big question..

[Diogo Almeida @CompleteSkeptic]

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI?

I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev

• 20-200x faster
• 40-400x cheaper (w/ output tokens free)
• Frontier composable intelligence optimized for decisions

AFAICT the shortest path to AI-based economic revolution

[Manoj Subramanyam @unscriptedmanoj]

@HermesWatcher Mobile via Hermes console but it has response interruption issue, not sure if it is of Hermes' or the console's

[Manoj Subramanyam @unscriptedmanoj]

@MattPRD @Muse Would love to try it out.

[Manoj Subramanyam @unscriptedmanoj]

@jackfriks He is truly an inspiration.

[Manoj Subramanyam @unscriptedmanoj]

@nahid_pro09 And no glm, only qwen 3.8 and deepseek are free.

[Manoj Subramanyam @unscriptedmanoj]

@nahid_pro09 Utterly useless. The context window is so limited that it is not taking anything other than Hi

[Alex Hormozi @AlexHormozi]

Three types of players:

People who play not to lose.
People who play to win.
People who play to play.

The people in the third bucket win the most.

[Manoj Subramanyam @unscriptedmanoj]

@HermesWatcher Hermes console by xpeta

https://t.co/5colgcQoqd

[Manoj Subramanyam @unscriptedmanoj]

@santiz01011 @thdxr The real comparison is when both are kept free.