xpo (@xponentcrisis) header
xpo (@xponentcrisis) profile picture
#00fa00
Joined August 2010
Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)Photo by xpo (@xponentcrisis)
xpo retweeted
Proud to be a day-1 provider on @Architect_Fi. Inference need real markets and we're bringing capacity to help build them. Congrats @BrettHarrison and team.
Today we’re launching Liquid Inference, an OpenAI/Anthropic-compatible router where inference providers compete on price for every prompt. Buyers route across 700+ open- and closed-weight models at the lowest available marginal cost, providers onboard instantly to start serving models.
10
38
6
143
24,814
We are thrilled to announce that @generaltensor has acquired @BackpropFinance and the Tensorplex Validator! Backprop Finance is one of Bittensor’s highest-volume trading platforms, facilitating billions in transaction volume since its inception a year ago. With a vertically integrated portfolio spanning from data centers to Bittensor subnets and validators, General Tensor is uniquely positioned to scale Backprop Finance and the Tensorplex Validator into the definitive global infrastructure for decentralised AI. We look forward to seeing these products thrive under General Tensor’s leadership.
1
11
548
Announcing the Dimensional Residency in Shenzhen. Deploy agents into the physical world. Real customers. Real deployments. We provide robots, $10,000 of LLM credits, free housing, office space, and customer intros to jumpstart your robotics company.
22
39
6
341
33,118
xpo retweeted
"Challenges the political economy of AI." That's how Jack Clark ( @jackclarkSF), co-founded @AnthropicAI, ran policy at @OpenAI, described what Templar is doing. He's now featured it in Import AI twice. 1/n
22
94
20
412
70,384
We needed to run trusted workloads on untrusted host machines. So over a year ago, we started building the Targon Virtual Machine to enable Confidential TEEs in production. Today we're sharing our white paper written alongside @intel: Decentralized Compute on Untrusted Hardware Using Intel® TDX and Encrypted CVMs
82
217
107
978
342,789
"It's like folding@home" Yes sir, welcome in. We’ve been waiting for you.
On the @theallinpod this week, @chamath asked @nvidia CEO Jensen Huang about decentralized AI training, calling our Covenant-72B run "a pretty crazy technical accomplishment." One correction: it's 72 billion parameters, not four. Trained permissionlessly across 70+ contributors on commodity internet. The largest model ever pre-trained on fully decentralized infrastructure. Jensen's answer is worth hearing too.
1
1
24
837
What if you could create an auto-research where your agent just focused on the eval and it was designed so that others could have swarms of agents across the web try to solve it and you paid them based on the ownership of the mechanism which produced the research
13
31
5
252
19,562
You know what is wild? Openrouter specifically ignores Chutes as a provider because, **checks note**: "they may use your prompts to train models" Reality: Chutes is the only provider which uses pure TEE and e2e encryption, private and cryptographically secure. You know who can read your prompts though? > Openrouter.
23
63
2
432
26,927
We have some big news to share today. Chutes is partnering with a research team from Harvard University to push the boundaries of AI inference efficiency. The team at Harvard, led by Professor Juncheng Yang @1a1a11a, is developing a new prefix caching algorithm designed to significantly accelerate inference while reducing hardware usage.
23
116
25
589
53,554
Unified, open, programmable, burstable, tradeable compute This is the way
Founder of Chicago-based prop trading firm DRW says compute will be world's top commodity in 10 years. People will spend more on GPUs than an oil, which means that there should be a futures financial market for GPUs.  Interesting implications for startups and cloud providers:
4
446
🏔️Introducing Ridgeline Behind the product is SN62 Ridges on Bittensor, where miners compete to build and improve the agent itself. Agents are continuously evaluated and scored, driving rapid iteration and capability improvements. Ridgeline is the product layer where those agents are put to work. • Assign a GitHub issue to the agent • It reads the repository for context and begins work • Generates and tests patches automatically • Your agent works in parallel to you, completing tasks without supervision • It’s like having another team member you can assign tickets to Inference runs across Bittensor infrastructure (e.g. @TargonCompute and @chutes_ai), while performance is measured on correctness, speed, and cost. Sign up now → shorturl.at/F0Tvt During the open beta, there’s no paywall, and new users receive 10 free credits to get started.
44
113
43
470
115,918
Using Claude Code or Codex? You don't need to pay Anthropic/OpenAI pricing for it. One command. Chutes becomes your inference backend. 60+ open source models. Same interface. Claude Code: curl -s -LO chutes.ai/chutes_claude_code…; bash ./chutes_claude_code_env.sh Codex: curl -s -LO chutes.ai/chutes_codex_env.s…; bash ./chutes_codex_env.sh That's it. Scripts handle the full setup, API endpoint, configuration, everything. Now your Claude Code runs on Kimi K2.5, GLM 5 or any of our 60+ models. Github: Claude Code: github.com/chutesai/claude-p… Codex: github.com/chutesai/response…
17
55
6
260
21,220
xpo retweeted
on the left: SAM3 on the right: @manakoai's base model with our new workflow, it takes us 15 minutes to go from left to right AND THEN we get miners to improve the base model, making it more accurate, faster, and cheaper via a fully automated manifest generating process i don't know about you but i love this game
12
20
4
137
16,076
Ramadan mubarak boys
Launching @Clawzempic today 😀 I spent $3k last month on inference for OpenClaw This month: less than $300! 💸 Same usage. Same models. Same lobster 🦞.. just sexier All it takes is 30 seconds: npx clawzempic🧵
3
468
The AI flies without accelerometers with pure visual inputs, oh man what else are they going to replace? 🤣
End-to-end neural networks racing drones, completely autonomous. Do you think what I think?
1
383
This feels like when reasoning was introduced into LLMs
them: engineering cannot be that interesting. how can you make people obsessed about your triple inverted pendulum? me: ~~ ♻️ Join the weekly robotics newsletter, and never miss any news → ziegler.substack.com
1
5
341
The first validator acquisition in Bittensor history, bringing a foundational piece of the network into our vertically integrated stack. RoundTable21 → General Tensor Validator
1
7
2
39
3,057
.@21RoundTable is now the General Tensor Validator (GTV). Same trusted operators. Same data center bunker. MORE conviction. It’s the same @wildsagelabs team that’s operated RT21 from day one.
2
8
2
28
3,363