CPA τalking τao - specialty accounting and tax support for the biττensor ecosysystem
Joined March 2011
- Tweets1.3K
- Following346
- Followers437
- Likes3.6K
τzork retweeted
The first of many coming demonstrators of why we built the IOTA SDK.
Importantly, this demonstrates three core use cases for us that iota previously didn't cover:
- fine-tuning
- distributed batch inferencing
- flexibility of the SDK - one of our engineers put this together very quickly, with all of the communication elements automatically resolved by the SDK.
More to come in the next few weeks as we get ready to launch!
Led by @v_truba based on Qwen3.5-4B and trained using ~879k decision examples.
But built different:
- Trained distributed, end to end, using streaming DiLoCo across 16 RTX A4000s.
- Training data came from our own stack: we generated and labelled it with distributed batch inference.
τzork retweeted
Everyone waking up to the idea of decentralized inference networks now, but also compute clusters, and training systems.
It will be protocols that mediate those networks since only a protocol can aggregate disparate and liquid compute -- across jurisdictions -- without taking on localized political risk.
gn datacenters --> gm inference networks
@engyai @chutes_ai @thinkinstant
τzork retweeted
Local inference isn't a terrible allocation of resources if you have technology that can unite distributed, heterogenous, and idle (re, off 95%) of the time compute to do frontier scale workloads.
That's the core of our thesis - and more broadly that maximising utilisation is the core next step in the AI race.
sam: local inference is a terrible allocation of scarce resources
“a lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of that”
τzork retweeted
were winning
sam: local inference is a terrible allocation of scarce resources
“a lot of people want local inference, and somehow they assume that the problem is solved now because we have open-source models, we're going to be able to run them on our computers, and so we're not going to need all the stuff in the data centre
but from a global allocation perspective, actually it's quite terrible to have everybody, like, have a bunch of the world's limited supply of DRAM sitting in their own computer and being used five percent of the time
you want maximum utilisation in the cloud if you're trying to deal with this very supply-constrained environment
somehow many people's intuition is the opposite of that”
τzork retweeted
EBITDA = Earnings Before I Trick Da Auditor
Anthropic has filed an S-1 registration statement with the SEC related to its proposed IPO.
Read more: anthropic.com/news/s1-sec
Legal disclaimer: A registration statement relating to these securities has been filed with the SEC but has not yet become effective. These securities may not be sold, nor may offers to buy be accepted, prior to the time the registration statement becomes effective. This post does not constitute an offer to sell or the solicitation of an offer to buy these securities, nor shall there be any sale of these securities in any state or jurisdiction in which such offer, solicitation or sale would be unlawful prior to registration or qualification under the securities laws of any such state or jurisdiction.
τzork retweeted
Actual is back. We migrated everything to bare metal. Service should be restored and a new update is available for your client which closes any upstream exposure to the massive amount of CVEs floating around here recently. One again, your computers were 100% safe + none of your info was ever public or leaked in any way. Sorry for the downtime but doing it this way was the only way. The team were like organized but feral wolves today
τzork retweeted
Actual Computer, which runs Bittensor's Subnet 95, says its new engine serves NVIDIA's Nemotron 3.5 Super VL at over 10,000 tokens per second. That is on a single NVIDIA DGX Station, across 256 simultaneous streams.
@opentensor
τzork retweeted
How Big Tech, ChatGPT And DeepSeek Could Lose To Decentralized AI via @Forbes
[hint: Bittensor $TAO]
forbes.com/sites/torconstant…
τzork retweeted
All of pearl mining is just spinning GPU matmuls, you can mine random matrices. So easily 99% of the hash rate is just burning expensive GPU hours with zero inference capacity. Maybe they can convert that into inference but it would be a completely new mining algorithm which acted more like toploc or grail (much more expensive to validate).
τzork retweeted
World’s most powerful magnet. Now in space. Almighty Magnet 🧲
We are very proud to announce:
The Almighty Magnet is officially in orbit!!
Launched this morning aboard @PortalSpaceSys’s Starburst-1 on the @SpaceX Bandwagon-5 mission, the Almighty Magnet is set to become the strongest magnet ever operated in space!
But we are truly just getting started.
The Almighty Magnet also carries a charged-particle sensor, measuring how our magnetic fields can deflect radiation. This data will take us a huge step closer to developing a new class of commercial radiation protection for electronics and biological life in space.
The future is electromagnetic.
τzork retweeted
I’m not sure which is more goated: our inference engine spitting out 10k tok/s or @jack15032’s graphic design
Actual Computer turns your NVIDIA DGX Station into a department-scale AI factory.
Read the full case study:
actual.inc/company/blog/case…
