I build humanistic technology for natural rights to prosperity, personhood, and privacy @humntech 🌱 I post on the intersection of neuro/crypto/philosophy.
if you’re automating anything, and you’re not using TAP, you’re missing out
>> full aws/azure/dns devops automation
>> full finance + bank loop engineering
>> scale policies securely to entire teams
@mbauwens@allmodelsRwrong@neynarxyz I read the announcement again, looks like they’re trying to be Singapore. Is the idea they ban anyone who ever posted impolite comments? Is it blanket ban on cuss words or is it idealogical?
pretty stoked to see this out, first of its kind afaik
now any private inference request can paired with an identity claim (is it a human? do they have social account X? are they on a sanctions list? etc.)
anthropic may require KYC eventually but there are other better ways
We’re proud to be using @humntech bridged USDC (cUSDC) to power private payments on Private Inference. 🛡️
Most AI services want your card, your email and your payment history. With cUSDC on @aztecnetwork private rail, you can fund your account without exposing any of that on chain.
What this means for you:
→ 💰 Top up with cUSDC. The sender, recipient and amount stay off chain.
→ 🚫 No card, no monthly plan, no separate accounts with each model supplier.
→ 🔑 One balance and one API key give you access to GPT and Claude models.
→ 📊 Each request is held at its maximum cost, then charged only for actual usage. Unused funds are released right away.
A big thank you to the @humntech team. Their work on bridged USDC is what makes private stablecoin payments for AI possible.
Network state conference was actually p good vibes and felt closer to the og idea. I went in with an intellectual apprehension to the landians I expected would be drawn to the flame but came away with renewed optimism. Who knows mb the whole thing is a precursor to a tech neofascist state, but I’m seeing that more and more as a sign of the times and less the desires of an elite
@leerob the blow off top of civilization ~ we’re boxing ourselves in, soon no one will know how to make anything truly new again
creative escape velocity to infinity
@CiaraNightingal Was great working with you Ciara! Best of luck with what comes next. I’ll let you know when the crypto swag Paris fashion show comes to fruition :p
cars are starting to pay for things on their own
@hebbianloop, in @Investingcom today, on what has to exist first:
"Give me hard spending limits, clear receipts, an instant off switch, and a named company responsible for resolving mistakes"
https://t.co/r7DbihA15Z
Zebrafish NN
It’s pretty amazing to learn that efforts to map the fruit FLY connectome (brain synapses) were successful and its amazing to see the digital fly brain successfully complete a whole series of tasks, including complex navigation of open worlds.
The fruit fly is 160,000 neurons and 10^7 synapses.
Teams are currently doing the same thing with a zebrafish which has a similar scale neuron count but 10x more synapses. Fly was 100TB raw data and Zebrafish is 200TB, so about 2x the raw data.
For context a rodent is 70m neurons and 10^11 synapses. Rodents will come later.
But this method looks like it creates totally outsized results with absolutely miniscule models.
Zebrafish NN is a vertebrae connectome, it has much of the same basic structure as other vertebrae, unlike FLY connectome.
We are just about used to LLMs and coding, but AI is really just beginning and there is a lot more, and a lot weirder stuff coming down the pipe.
These connectomes have already shown to be incredibly resilient, you can blind their sensors and injure their outputs and they still succeed.
They are incredibly compact.
A whole lot of inanimate objects are going to get complete autonomy, totally offline, air gapped autonomy. They will be delivering pizzas and fighting wars.
Connectomics is an opposite approach to LLMs but you can of course use LLMs to help develop connectomes.
It’s amazing that you can take a biologically evolved brain, map it, produce a digital twin, and then run it at machine speed.
Imagine a human brain accelerated from 100Hz neuron fires to 3,200,000,000Hz that a typical CPU runs at. That’s 30 million times faster. But would need 1-2 exabytes to map.
Anyway “Zebrafish” is next.
I wireheaded the fly and forced it to doomscroll flytok.
Dopamine neurons are measured and artificially enhanced to ensure maximum enjoyment.
My goal is to create a fly that is happier than all other flies combined.
you really nailed it, the analysis is spot on and generalizes well to agents
i think most builders will read this and set about making the perfect infra
the more savvy approach imho is literal sandboxing of capacity ~ tools like https://t.co/79nw5qRili or paradigms centaur that make acting out of requirements impossible by design