Everyone needs an AI brain, these two lay out. Theirs caught my attention because their database, or "brain," has a subconscious. In other words, over time, the brain itself stores things in long-term memory better than anything else I have been shown so far. This has deep implications for... everything we are about to do with AI. Even when getting into an autonomous car, if it knows more about you, it will serve you better. These two, cofounder/CEO, Prakshal Jain, and Alex Morris cofounder laid out why their approach will: • Hallucinate a lot less, if at all • Be faster to run • Be cheaper to run This works whether you have a personal brain on your own account, or a company brain that watches your business work by rolling up context from each individual machine. While this is very scary new technology for people, the sheer cost advantage is what really got my attention when visiting this company in San Francisco: Even though we will argue about the privacy implications, which we cover at depth in this interview. Here is the future of agentic systems.show more

They call it the graph of everything you know.

Here's what Grok learned by watching this video.

For more information, follow @MitosisLabs, me, or @prakshaljain_

Company brains are great but being able to access and apply context to what your working on can drive much more roi.

They don't disagree

I like it, tried signing up with my google account but it failed twice.

btw company handle: @MitosisLabs if you wanna stay tuned! Thanks :)

You were already on my lists.

I have not invested time in brains other than a folder with Claude or agents md, subfolders, and md files, per project, projects sitting in a master folder. I think the models themselves are going to eat all this activity in 6-8 months.

Crazy

Welcome to the future!

long term memory is the only thing that separates a cool demo from a tool u actually use everyday

The autonomous car example makes the privacy tradeoff compelling. How much history is too much?

Whole pitch rests on the word subconscious, and that's the bit nobody's defined.

The cost advantage plus better recall is a rare combination.

This is awesome

redirect threw a 502.

I like what these guys are doing. Wait until you see what I'm doing with my next gen memory system, REkall. Very similar, but it goes much deeper. Soon...

ai with a subconscious is wild ngl

😉🆙⬆️👆

subconscious memory changes the game

long-term memory at the infrastructure level changes everything for agents.

Will give it a look later, but sounds like nothing more than custom guardrails with a long context window. *more vectors to process doesn't prevent hallucination, per se, but narrows the space for potential f-ups.

long term memory sounds cool until it remembers the wrong stuff forever.

yeah but who decides what goes in it

"rolling up context"

Impressive vision where does privacy fit into this?

If you run the models locally and you store your data locally, then you have control. If you use a cloud service like OpenAI, xAI, Anthropic, or Google Gemini, they are studying you to train their models. You're giving them what they need to get to AGI, and this is a monster that will just keep sucking up data if you let it. You have to choose your architecture wisely. I think these two guys lay it out well. If you watch the video, you'll understand a lot about what you should be thinking about when it comes to memory systems or AI brains.

The idea of an AI "subconscious" long‑term memory sounds really fascinating.

Eventually AI will know what you want before you consciously know what you want. Maybe so much so that you can end up in a weird AI induced psychedelic experience. Like when the recommendation algorithm funnels you down a rabbit hole, but as immersive as real life.

Mine