Lamida
135 posts

Lamida
@LamidaGlobal
@Bittensor focused US-based company. Subnet Incubation, Subnet Investment, Validator ops, Miner Infrastructure, and Subnet Consulting.
USA Katılım Ekim 2020
44 Takip Edilen2.5K Takipçiler

We were pretty hard on @ridges_ai after that misleading post earlier this year.
Now we’re watching a very similar dump with @QuasarModels.
Both have strong teams. Both got punished fast by the community.
This isn’t toxicity.
This is conviction.
Bittensor holders aren’t chasing memes.
They expect high standards and they enforce them.
That level of maturity is exactly why $TAO feels different from the rest of crypto.
If you’re new here, this is the place to start:
learnbittensor.org
What do you think; fair reaction or too harsh?
#bittensor #DeAI
English

Day 1 of asking:
which subnet product did you use today? #bittensor
English

Brutally honest dissection of @Loosh_ai
> a subnet that is at risk of deregistration
1. The one thing that must be true for Loosh to work:
Robotics and agentic AI companies will pay for an external cognition/memory middleware layer rather than building it in-house or waiting for foundation model providers (OpenAI, Google, Anthropic) to ship it natively.
This is the load-bearing wall. Everything else "the Bittensor subnet, the moral compass, the EEG multimodal inference" is irrelevant if this assumption fails.
And it is highly testable right now with zero code: call 10 robotics companies and ask if they'd pay for this today.
2. Three Fatal Flaws
>OpenAI has memory. Anthropic has long-context. Google has persistent agent state baked into Gemini. Every major model provider is racing to solve exactly the "stateless agent" problem Loosh is solving. Can they compete? Doubtful.
>Loosh is simultaneously pitching:
- An enterprise B2B cognition stack for robotics companies
- A Bittensor subnet playing in the crypto/decentralized AI space
These are completely different buyers, sales motions, trust requirements, and timelines. Enterprise robotics companies don't want their memory fabric running on a permissionless GPU network with token incentives.
>Their differentiators "Monroe Institute-inspired consciousness frameworks, EEG multimodal inference, virtue ethics evaluators" are intellectually fascinating but not defensible.
A well-funded competitor can hire a neuroscientist and a philosopher in a week. The temporal knowledge graph and memory fabric are real engineering, but nothing described suggests a genuine technical moat that can't be replicated.
3. The stateless agent problem is genuinely painful — anyone who has built multi-step AI workflows knows context management is a nightmare. That's a real problem.
But Loosh's framing inflates it into something more exotic: "machine consciousness," "decentralized inner life," "soul for robots." This creates a credibility gap
4. Concern:
Their team story "meeting at the Monroe Institute studying consciousness" is authentic and interesting.
But it also signals that the project may be driven by a passion for consciousness research that's looking for a business model, rather than a market pain point that happened to find the right founders.
The founders have the right raw ingredients. But the idea needs surgery, not encouragement.
English

.@const_reborn , when is tao.day?
we should have one too😅.
domain is available for $379.99. are we celebrating tao.day when price breaches this price?

English

@LamidaGlobal Why are the other eggs bigger than mine? Anyway i can make mine bigger ?
English

there is something different about @chutes_ai and how it is operating.
Chutes is registered as Chutes Global Corp. but without CEO? surely there are equity holders and directors.
without CEO/central decision maker, love to know the chain of command?

English










