Jemmy (τ) (@Jemmy_AI) header
Jemmy (τ) (@Jemmy_AI) profile picture

AI/ML, Finance, and Bittensor Enthusiast. 50% $BTC and 50% $TAO is the only way

Joined June 2020
Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)Photo by Jemmy (τ) (@Jemmy_AI)
send pearl to 0
2
71
what if the fed just keeps faking high inflation for the next several years so they can raise interest rates to 10%++ great reset
1
52
show them @webuildscore
This is why computer vision is hard
Readers added context they thought people might want to know
The sprinkler activated due to heat from a real charcoal grill stove behind the sign at a barbecue stall (a customer was grilling skewers), not the flame graphic on it; the resort confirmed normal operation and adjusted the stove angle. 163.com/dy/article/LC4… k.sina.cn/article_532885…
64
send noid to 0
1
136
Jemmy (τ) retweeted
The new weekly competition on SN 111 Claims is live! 🧐 What makes the competition compelling and why would one care? In Claims’ new competition, miners submit their whole approach and Claims evaluates which pipeline does the best job before putting the winner into production. 📜📜 That could reduce duplicated work and let the subnet turn more emissions into structured scientific data. 🧑‍🔬🧑‍🔬 The goal is spend less of the subnet's emissions paying for duplicate paper processing and more on improving the extraction system. 📜📜 Claims estimate this could produce 10-100 times more paper coverage per dollar of emissions. 📈📈 I will be keeping an eye on this "claim" to see if it can blossom into fruition. 🍎🍎 @DeSciClaims teapawt.org/DeSciClaims/status/211…
Our new competition is live! 31 submissions are competing. We'll run the winner in production next week. 🚀
2
3
66
Scientific research produces an enormous amount of information, but making sense of all those papers is still expensive Around $130B is spent every year on researchers reading and reviewing scientific literature, out of nearly $3T in global R&D spending As AI agents take on more of the evidence synthesis work, the bottleneck starts shifting from reading papers to having reliable, structured data they can actually reason over @DeSciClaims is building a data layer for biotech, pharma, and research agents to work across tens of thousands of scientific papers The opportunity is estimated at $30B. Pretty big problem to solve Whitepaper: claims111.ai/whitepaper
Our new competition is live! 31 submissions are competing. We'll run the winner in production next week. 🚀
1
3
130
there is a better $PEARL but its still at $13 million market cap still accumulating...
109
why cant we have a $TAO discord server tag, i want to show it off to my discord friends🙂‍↔️
1
119
Every liq on @ForgeLending becoming fresh data for Endure Network (the subnet), miners compete to develop better risk parameters based on actual data over 50% of protocol revenue will toward SN30 alpha buybacks @webuildscore is the first collateral about $4mill in TVL commit Curious to see how this performs once Forge goes live on mainnet
Heat’s rising. Bittensor is about to get liquid. Mid-November.
2
2
93
97% of miner-submitted claims were approved by late October, up from 80% when the judge panel first went live this is the progress @DeSciClaims has made with its open competition model. Independent miners submit claims every week and compete against a reference pipeline, with blind evaluations from dual LLMs there's another interesting detail... when miners disagree with the reference pipeline, judges now side with the miners 3 out of 4 times the reference pipeline isnt automatically treated as the answer. Miners can challenge it, and their work has to stand up to independent review this is a compelling way to build a scientific dataset
Our new competition is live! 31 submissions are competing. We'll run the winner in production next week. 🚀
69
most of the conversation around Alpha lending is about borrowing... but where does the $TAO liquidity come from? @ForgeLending is trying to make depositing $TAO means earning root yield as a base, with additional interest coming from borrowers as utilization increases borrowers also get something out of it as well, their alpha collateral keeps earning native emissions, with 75% of those emissions going back to them. both sides have a reason to participate. The testnet is live, so worth trying the flow before mainnet
Heat’s rising. Bittensor is about to get liquid. Mid-November.
1
5
179
thank you for still being with me @Bitcast_network
2
39
your government decides where the printed money goes in bittensor, you are the one who decides which subnets the printed $TAO goes to the decentralization of economics is here
16
421
bearish on tao is a bullish catalyst
7
175
not enough pain, send it to $5k
last dump before we go to $10k ... im talking about bitcoin
1
8
168
and he sacrifice, THE OWNER EMISSIONS!!!
⚖️ Emission split changed SN51 lium․io owner 18% → 9% · validators 41% → 41% · miners 41% → 50% #Bittensor $TAO
4
183
Jemmy (τ) retweeted
The biggest liquidation on Mentat Lend so far happened yesterday, and the liquidated borrower still holds 94% of his value. Here is how:
This quoted post is unavailable.
3
3
28
616
4 GB was an annoying limit for running models in the browser. Most useful AI model simply wouldnt fit in a chrome tab @TPN_Labs got around that by forking the inference runtime and enabling 64-bit memory addressing, pushing the limit up to 16 GB So you can actually run real AI models directly in a browser tab now, without installing a separate app If you want to try it, Gratis works on Chrome, Edge, Firefox and Brave gratis-ai.com
We're excited to introduce @Gratis_AI. Run opensource AI models locally on your machine for free, with no setup required. Gratis runs models from @huggingface directly in a browser tab or desktop app for more power, making every chat private by design. TPN builds small, capable models; gratis runs them. gratis-ai.com
8
209
Less than a month ago at @SubnetSummerTAO wrote about how prompt-to-print would make a great subnet (by pure coincidence). Bittensor's incentive mechanism is perfect for this: jobs rotate through miners, every design goes through engineering validation, and miners who miss the deadline lose the job to the next in line. @CAD_mind_subnet is essentially the first IRL-focused subnet. This is huge for engineers and hobbyists too. Engineers can skip hours of CAD grunt work and get a validated first draft, with STEP files, parts list and assembly instructions, in the time it takes to write a brief. Hobbyists with a 3D printer and an idea no longer need to spend months learning CAD. Describe it, print it, build it. Imagine them building a drone for @SwarmSubnet or a robot head for @openroboto. Subnets designing hardware for other subnets. 3D printing is now live on Bittensor. ⚙️
CADmind is LIVE on Bittensor. 🧠 Prompt to print. Create custom functional designs with natural language prompts. From parts and PC cases to drones and robotics. Where Bittensor intelligence meets the physical world. Backed by @YumaGroup 🌐 cadmind.ai 🧵 1/5
3
2
1
24
1,161