- Tweets 4,065
- Following 891
- Followers 299
- Likes 29,042
Hard disagree.
True decentralization is a death sentence if the protocol hasnt been iterated on and battle tested.
We would never have arrived at dTao had bittensor been decentralized too early
Let Const cook $TAO
$TAO as an experiment has failed, there is no doubt. But as a proof of concept it did well. I expect a competitor is already in the works. True decentralization, with no CEO like Const. Bittensor cannot be saved but deAI can
2
144
SN95 Actual $TAO subnet makes sense to me, if they
a) reward Actual client users/inference providers with alpha tokens, to buil up a pool,
b) sell that pooled inference for profit,
c) prop up the value of alpha thru buybacks
OR...
1
1
89
What competitive advantage do they gain by having continuously improving inference providers (improving on what metric?)
Subnet 95 does not make sense to me as a mere ledger + vehicle for user subsidies
Are Actual going to incentivize cluster size? Stability? Performance?
1
8
...designed to make stranded resources competitive.
Subnets are supposed to incentivize things that can be optimized through innovation.
The key question for me is:
What can Actual optimize for by having a subnet, that will make their product better
1
Marforio retweeted
Super impressive presentation from Gareth at Exploit on VidAIo (AI video compression).
Video files are HALF the size of AWS compression.
Could literally chop @netflix's $1B per year streaming bill in HALF.
powered by Bittensor
Full vid
stream.vidaio.io/vod.html?ta…
10
12
60
2,718
Marforio retweeted
There was 366tb stored on @Hippius_cloud a month ago.
It is currently at 887tb. Just a casual 142% MoM increase.
Now with confidential computing, @Hippius_cloud is no longer just a storage layer. It's somewhere you can store, ship and run.
bittensor:native
9
17
101
7,812
We’re not here for the churn and burn model that defines much of the prop firm industry. We’re traders, and we built this for traders. We wanted a prop firm with open-source rules, on-chain payouts, and decentralized infrastructure that aligned incentives with traders. We wanted to build infrastructure traders could trust, rather than ask them to trust our words. Most importantly, we wanted that infrastructure to put traders first.
Our next steps are focused on setting traders up for success, helping them prioritize risk-adjusted returns, and giving them the infra to do it.
We’re giving traders the option to skip evaluations entirely and choose how they scale into Pro accounts on 10x the capital, all powered by decentralized, onchain infrastructure. Something that’s never been done before.
Then, we’re releasing the infrastructure for traders to develop, test, and refine strategies using AI, giving them the tools to succeed and scale on Pro accounts. We’ll continue to put traders first and build the next generation of infrastructure around their success.
Our goal is to identify great traders, give them the tools and capital to succeed, and copy trade their performance. That's where we want to build our business, not around how many traders we can push through evaluations.
The prop industry has spent years building systems around trader failure. We're building one around trader success, and we're willing to challenge the entire industry to do it.
Instant funded accounts: LIVE NOW on Vanta! 🚨
• Only 100 available
• $25k account - no evaluation
• Realize profits, get paid weekly
Even better? They're on sale. Use code OCT20 for 20% off.
Get one before they're gone 👇
vantatrading.io/instant-fund…
4
9
35
2,328
Marforio retweeted
What costs over $3,000 a month on AWS..
@Hippius_cloud does for $245
93% less for 50 TB of storage
Another example of a single #bittensor subnet completely changing the game
$TAO
8
44
249
10,260
Marforio retweeted
Carbon is live! Subnet 116.
Physics AI is now on Bittensor
Let’s go!!!
13
24
121
21,160
Marforio retweeted
Programmable bandwidth only works if the floor is cheap.
Beam base price: $5/TB delivered. AWS egress: ~$90/TB.
Same job, ~18x cheaper, with routing you can actually control.
(Powered by Bittensor)
4
28
95
12,808
I’ve been asked a few times now and I’ll be honest, you won’t see me at many events. I’m not here to socialize. I have 2 little kids and my life is mostly working all day and then spending what remaining time I have with them. As my team knows, I work practically 24/7. My passion is Vanta, I have no other pastimes. All I care about is disrupting prop trading and we’re well on our way there.
10
3
58
4,614
Marforio retweeted
Step 1 - Build useful product
Step 2 - Win real customers
Step 3 - Generate external revenue
Step 4 - Use some of that value to support/burn Alpha
Result = Successful product creates economic value for its own subnet
This is the playbook that #bittensor subnets should be chasing
Good work @webuildscore 👏
$TAO
September was the first month we broke even.
Two private track partners are now long-term paid clients.
We’re converting clients faster than before, still in sports, and now in fuel retail and security via Manako. Details will be shared in separate posts.
That was the signal we were waiting for. Buybacks (and burn this time) of sn44 have started from our owner address.
First ones are in: 9 × 1,044 and 1 × 4,444. Not from Manako. Purely from the subnet.
We will not talk about them. No schedules, no amounts, no marketing. A buyback today says nothing about tomorrow. It could be $1,044 on day 1 and $4,444 on day 27. Or neither.
They will happen.
They will not be predictable.
Trust the process. Trust the vision.
13
22
180
11,494
Marforio retweeted
At our Exploit Summit keynote, we showed the next phase of Minos: a native macOS miner app where AI agents autonomously run genomic research loops to turn research questions into eval tasks, with an iPhone companion app that lets you operate it from anywhere.
We are very excited about what’s coming next. Stay tuned.
A Mac app for miners. An iPhone companion to control it remotely.
“Let your agent do the work for you.”
@centrum_blue previews @theminos_ai’s second incentive mechanism: AI agents working on genomic research tasks for Bittensor’s SN107. Watch the on-stage demo run a research loop, open its logs and pause the miner from a phone.
1
14
58
3,373
Marforio retweeted
Subnets using Subnets.
Powered by Bittensor.
GM is our 'openrouter'.
Beam is our 'Big Bandwidth' provider. Move massive files fast and securely with granular enterprise-class controls.
Beam Studio is live and we're proud to be a launch partner.
Every prompt in Beam AI, from building workflows to investigating failed runs, is routed through SayGm.
@b1m_ai said "we could have picked any provider but SayGM is the best".
Congrats on the launch.
7
19
128
5,439
Marforio retweeted
Engy is getting faster.
Pareton now tunes how @engyai runs its models. Quicker replies. Same API on your side.
Engy is one of the fastest-growing inference networks on @bittensor and beyond. A big customer with serious traffic. That is the workload we were built for.
7
27
95
15,632