Krishna©️Thipparthy

Non-Local Biological Engine in ©️onsciousness | Building @MetarealityTech

[Krishna©️Thipparthy @KrishChaiTech]

The closed source era ends,
The open source era begins

[Krishna©️Thipparthy @KrishChaiTech]

Not your weights not your model

[Krishna©️Thipparthy @KrishChaiTech]

Elon Musk’s net worth just broke physics.
One man went from rockets to robots to brain chips and somehow bent reality itself. We’re 100% in the simulation — and Elon just punched a hole straight through the code.
Now he’s calling it: we’ve entered the Singularity, with 2026 as the tipping point. The glitch is real. The future is accelerating.
What a timeline. 🚀🕳️🤖

[Krishna©️Thipparthy @KrishChaiTech]

@alokbishoyi97 @EVO__HQ Good direction, evo beats sia

[Krishna©️Thipparthy @KrishChaiTech]

@naval @maxhodak_ It’s like trying to measure something without knowing what it actually is.

[Krishna©️Thipparthy @KrishChaiTech]

We are heading into a digital amnesia & cultural Alzheimer's with data centres handling the mind control mechanism

[Krishna©️Thipparthy @KrishChaiTech]

Silicon = Neuron

[Krishna©️Thipparthy @KrishChaiTech]

@bevedoni @karpathy Maybe he thinks it’s already safe

[Krishna©️Thipparthy @KrishChaiTech]

Indian Gen Z’s have take over the political CYBERSPACE 🫡 every political party must follow

[Krishna©️Thipparthy @KrishChaiTech]

@runfusion Can you explain the best way to pair paperclip and runfusion together ? Immature request but would be happy to know how you guys workflow looks

[Krishna©️Thipparthy @KrishChaiTech]

VRAM VRAM VRAM is all the buzz

[Krishna©️Thipparthy @KrishChaiTech]

@pmarca Physics is all we need 😉

[Krishna©️Thipparthy @KrishChaiTech]

@bindureddy We need hallucination reduction and a native memory system for the agents, not the human readable memory systems, something more fundamental and hallucinations needs physics based solutions inside the black box 📦

[Krishna©️Thipparthy @KrishChaiTech]

@GoodfireAI @thoughtson_tech Please check dm, pls share access to silico

[Krishna©️Thipparthy @KrishChaiTech]

@runfusion @aronprins I have a question let’s say I have a team of 8 agents in paperclip, can I pull them into fusion and coordinate them from both places ?

[Krishna©️Thipparthy @KrishChaiTech]

Very true, it feels like intelligence needs some kind of a grounding to reality also the human readable memory layer we built around the llm is accelerating what you actually said … agents need their own memory that is not human readable. This helps 50% I guess rest of the 50% is still hallucination inside the llm black box itself ..

[Krishna©️Thipparthy @KrishChaiTech]

@steipete 😂😂 share it with us pete

[Krishna©️Thipparthy @KrishChaiTech]

Founders from the past : How many employees do you have ?

Founders from the future : How many agents do you run ?

[Krishna©️Thipparthy @KrishChaiTech]

Alright we are moving away from transformers. 12M context is nuts 🤯🤯🤯

[Alexander Whedon @alex_whedon]

Introducing SubQ - a major breakthrough in LLM intelligence.

It is the first model built on a fully sub-quadratic sparse-attention architecture (SSA),

And the first frontier model with a 12 million token context window which is:

- 52x faster than FlashAttention at 1MM tokens
- Less than 5% the cost of Opus

Transformer-based LLMs waste compute by processing every possible relationship between words (standard attention).

Only a small fraction actually matter.

@subquadratic finds and focuses only on the ones that do.

That's nearly 1,000x less compute and a new way for LLMs to scale.