Hash Rate - Ep. 179: Chutes PARALLAX
🧙 Guest: @jon_durbin of @chutes_ai
Bittensor $TAO Decentralized Training on SN64
0:00 Intro to Chutes
1:10 Chutes SN64 By The Numbers
8:32 Parallax Deep Dive
51:34 Parallax Business Model
54:22 Audience Questions
I am in and using this now.
OMG is it well-done.
The UI is VERY clean and extremely usable.
Pay in TAO, USDC, subnet tokens or cc.
This is going to do a LOT for Bittensor.
We just let the first 50 on the waitlist in.
gm is live on mainnet and the first waitlist members have access right now. Check your inbox.
Every account got $10 of credit on us. Go break something.
pushed a big autofrontier 'recursive self-improvement' update tonight.
the loop now improves itself at three levels:
level 1, it already did:
train better models overnight, a strict gate deciding which gains are real.
level 2, new:
the gate also tests the strategy used to find those gains. better strategies kept, worse ones dropped.
level 3, new:
the loop can improve its own code, same gate judging.
so one night of GPU gets you a better model AND a better way of finding the next one.
every night compounds.
repo in comments.
Our top miner's voice model is closing in on SOTA performance - at a fraction of the cost.
Vocence already has bunch of paid users.
ElevenLabs' moat is about to disappear.
1000 TAO buyback is coming soon.
And that's just the beginning.
Stay tuned.
#Bittensor $TAO
$2,200,000,000... that's how large the top AI Crypto is right now
While I believe that's undervalued, I think within that are 1000x plays that are abundantly obvious
$TAO for example
>why would I buy TAO when I know a majority of subnets are just emission farming
>size into subnets with clear revenue or are likely to produce success stories right now (chutes, vanta, targon, iota, minos, nova that are likely to carry the network to new highs
If they're successful you might catch a 10x on TAO, but you will catch 100-1000x on subnets + farm yield from dTAO
There's almost no reason to be allocated directly into TAO alone unless you're lazy or believe subnets won't grow
Which in that case you shouldn't be holding TAO at all
We are aware of the security incidents affecting both the ORO and Connito subnet owner accounts and are actively supporting the teams as they investigate the circumstances surrounding the compromises.
Based on the evidence available today, we have found no indication that the Crucible wallet, its infrastructure, or its security architecture played a role in these incidents.
Our teams continue to work alongside the affected subnet teams to determine the full scope of the incidents, identify the root cause, and implement any additional security measures that may be appropriate.
The security of our products and the broader Bittensor ecosystem remains our highest priority. We are committed to sharing updates as we verify new information.
We appreciate the community's support while this work continues.
Here's the design behind VocenceBench. All technical details are verified through real model testing.
Read the full paper - feedback will be really appreciated
https://t.co/zDJP5AYOmu
We're cooking. Nonstop.
Product idea: as soon as I say "Bye" on a Zoom call, it should disconnect me. Else there's always those 1-5 seconds where both people are looking around for the "leave" button. Awkward. This would be the next quadrillion dollar company
Probably related somehow.
And probably a key management issue in some way.
We don't know much just yet -- other than this does NOT appear to be a chain-level thing.
Everything we have built at 404 has been powered by Bittensor’s decentralized network and our commitment to pushing 3D asset generation forward.
We are here to keep building on that foundation for the long term, and we are ready to keep proving that commitment to the community.
Since June, 404 has been convicting 100% of new emissions on SN17.
Today, our owner lock sits at around 48.6k Alpha, with 100% Conviction and perpetual lock enabled.
This new Conviction update makes locked positions, unlocking activity, and long-term subnet commitments visible and verifiable on-chain.
For us, that matters because commitment should not only be stated. It should be proven.
We are not building on Bittensor for short-term incentives. We are building for the long term, and our Conviction now makes that commitment clear on-chain.
The @mentatminds team is exactly right, and it's why we built the Yuma Composite Index, YCX.
Sum of Subnets is an unweighted total of Bittensor subnets: any single subnet can move it. Over the past 30 days SoS rose 7.5% while YCX rose only 0.6%.
The gap is almost entirely two subnets: SN40 (0.16% of network market cap) up 555%, and SN116 (0.39% of market cap) up 253%, both moving on thin liquidity.
YCX is market-cap weighted so no single illiquid subnet can swing the whole index, the same way a microcap tripling doesn't move the NASDAQ composite.