τp (@tpwereld) profile picture

Trading enthusiast and AI explorer on Bittensor | sharing what’s on my mind

Amsterdam, The Netherlands
Joined March 2024
Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)Photo by τp (@tpwereld)
Pinned Tweet
Ikigai and Bittensor - a point of view I have been thinking about Ikigai lately and how surprisingly well it maps onto #Bittensor and $TAO I know the Ikigai concept from my professional career and it’s often visualized as the intersection of four things: 1. What you love 2. What you are good at 3. What the world needs 4. What you can be rewarded for Your “reason for being” sits somewhere in the middle and I think Bittensor subnets are going through a similar discovery process. In the early dTAO days, simply building something was often enough. Launch a subnet. Create an incentive mechanism (IM). Attract miners and validators. Get emissions. Simple. But as we mature, I think that is changing and the ones adapting, are the ones surviving. The subnets that ultimately matter, will find their own version of Ikigai. Recognising subnets that may have found their Ikigai, is a way to value subnets, incremental to just on-chain metrics and GitHubs. Let me talk you through an example that comes to mind: 1. What you love → Founder conviction The best teams probably will not be the ones chasing whatever narrative currently attracts emissions. The best teams are obsessed with the problem they want to be solving. Models. Inference, Math. Agents. Data. Compute. Whatever it is. That obsession matters because building something genuinely differentiated takes months and/or years, with true passion. 2. What you are good at → Competitive advantage Passion alone obviously is not enough. A subnet needs to produce something better, cheaper, faster or fundamentally different. Your competitive advantage has to survive competition, in and outside Bittensor. 3. What the world needs → Product-market fit This might be the biggest filter and qualification. Does anyone actually need what the subnet produces? Not validators, not alpha/tao holders, but actual users. A subnet can have a beautiful incentive design, beating any SOTA benchmarks, run sophisticated ML loops, and maintain a passionate community, but if nobody ultimately wants the output, what exactly are we incentivizing? 4. What you can be rewarded for → Sustainable economics And finally, economics. Bittensor adds something unusual to the Ikigai framework. The network itself can bootstrap the search. Emissions provide economic energy to miners, validators and builders before traditional product-market fit necessarily exists or is proven. But I don’t think emissions are the destination. They are the runway to profitability at scale. This is something that we truly should learn from corporates: profits at scale. Over time, the strongest subnets should increasingly be able to demonstrate why the intelligence they produce have economic value. And this is where I think the concept of Ikigai and Bittensor gets very interesting. Because these four circles don’t automatically overlap. You can have great technology nobody needs. Huge emissions without meaningful output. Strong demand but weak incentive design. Or passionate founders without a real competitive advantage. The market/staker responsibility is ultimately to figure out which subnets sit closest to the middle. And perhaps that is what Bittensor itself is becoming: A giant decentralized search for the Ikigai of machine intelligence. Thousands of people allocating capital, compute and intelligence toward different hypotheses about what AI should produce. Most will probably fail. Some will find temporary economic games. But a few may find that rare intersection where: they love building it, they are uniquely good at it, the world actually needs it, and the economics sustainably reward it. Those are the subnets I want to find. Finding the Ikigai in Bittensor.
1
3
549
My entry for the @xeveryframe Wildlife Documentary Competition, made on EveryFrame 🎬🐾 teapawt.org/xeveryframe/status/212…
EveryFrame AI Film Contest No. 01 is open 🎬 Wildlife Documentary Competition: a 2-min narrated wildlife film, every shot on EveryFrame. 🏆 $5,000 Prize value to one winner 🗳 Community votes 📅 Closes 27 Nov, 12:00 UTC Make it: app.everyframe.studio Enter 👇 everyframe.studio/contests/w…
5
5
2
21
1,850
Crypto still the best industry to be in. #crypto
5
233
Small reminder for the crypto boys out there
1
42
All liquidated crypto traders today that are watching the market unfold now #crypto #feelingalive
Villains watching Goku vs Kid Buu in Hell — Episode 280, Dragon Ball Z
3
603
Decentralized AI is a necessary counterweight to regulatory capture that shields AI oligopolies from competition. Bittensor gives people a way to build that alternative and own a stake in it.
A solution like bittensor:native has to exist where we can incentivise (AI) workloads in a permisionless way With Jensen out i have very little trust in Sam/Dario to do the right thing, they just want to have full monopoly and power concentration Bitcoin was spawned against the banks, bittensor can be that against ai nationalisation.
15
82
5
492
22,551
dTAO market is like a pawn shop now. If you dig and research deep, you find many undervalued subnets. Subnets with cracked owners, Gud ideas, and crazy backing, at bottom tier prices. This is the time to stack subnets and support those subnets. bittensor:native
3
11
728
A solution like bittensor:native has to exist where we can incentivise (AI) workloads in a permisionless way With Jensen out i have very little trust in Sam/Dario to do the right thing, they just want to have full monopoly and power concentration Bitcoin was spawned against the banks, bittensor can be that against ai nationalisation.
29
123
9
689
48,296
I've spent the past few weeks reading 100s of public data sources about AI development. I now believe that recursive self-improvement has a 60% chance of happening by the end of 2028. In other words, AI systems might soon be capable of building themselves.
285
495
236
3,496
1,676,019
Visionary
We will turn the web into a brain. We will decentralize. We will be permissionless. We will be free.
1
109
Expect all new/pivoted subnets to massively focus on revenue and bringing value back to the alpha token. That's the narrative. There are some incredibly undervalued plays in dTAO now. They are obviously not the ones shouting the narrative. Those are already priced in. $TAO
3
6
611
What will the world's marketplace for intelligence be worth? $TAO: The Commanding Heights of Intelligence is our case for who owns the economy once intelligence gets cheap @jason @markjeffrey @rob_svrn
32
61
19
218
33,057
We have to be careful in our bull posts though as many are starting to point sticks at the difference between revenue and profit - I know, they are never satisfied and can just be people looking for something to say for the sake of something to say ...but we should also be considering and starting to consider our positions there too. Also how much of this is provable on-chain - like end to end provable. And if not. what are our responses to that.
6
1
22
1,181
🤖 AI’s next battle may be over who controls the compute. Bittensor co-founder @const_reborn tells @2084Richardson at @ExploitSummit why open-source AI could threaten the biggest labs. The interesting part starts when they notice. 👀
21
141
30
489
93,521
INSIGHT: #TAO is trending after unveiling quantum-resistant wallets, giving users post-quantum signatures and the option to replace signing keys automatically.
80
207
32
1,195
68,094
He bought a subnet. He launched it. He bought its alpha before anyone else, in the same block. He sold everything within 15 minutes. Then he registered a single miner, his own, and has been collecting 90 % of the emission ever since. 19 days. 539 τ net profit, paid by the 211 other buyers. The subnet is for sale. Fifteen days later, he did it again on another subnet. With the same wallets. We read the chain, block by block. Full dossier tomorrow, here and on our Telegram: t.me/tensiafoundation Clarity is Alpha.
11
5
4
68
13,183
Good perspective
Some investment discipline needs to return to #Bittensor. Customer revenue should cover costs and reduce subsidy dependence before funding discretionary buybacks. I would give more weight to customer-funded burns from lightly subsidised subnets than to large burns from the biggest emission recipients. If development still needs investors’ support, why divert available cash to supporting the token price? Is it solely to increase the token price to get either more free money? The price already include an expectation for the future subsidies. Why to amplify it even more? This is a vicious circle. Return unnecessary support first, buy back and burn from a genuine commercial surplus afterwards.
1
81
Replying to @rousseaujp
Apologies, I forgot to send! Here are today’s numbers (Oct 29): 817,560 memories — approaching 1M. Since Oct 11: • +185,684 memories (+29.4%) • +95 registered accounts (+3.5%) • ~10,076 net memories/day on average heyditto.ai/dashboard
🤖 Made with AI
1
1
38
coding agents were the first agents to really make an impact. company agents are now on the horizon, and they come with some different requirements and primitives: * ux - agents live where you work (slack) * identity - agents can act obo the user or with a fixed identity * memory - agents remember preferences, feedback across sessions * sandboxes - agents can write code in a secure way * schedules - agent runs can be scheduled (by users and by agents)
17
16
138
11,187
Lmao
De pitch van de @VVD (@QueenyRaj) om miljardenbedrijven naar Nederland te halen: “Kom hiernaartoe, dan kunnen we jullie belasten en precies vertellen wat jullie wel en niet met AI mogen doen!” Totaal absurd. Geen deepfake. We worden geregeerd door clowns.
69
We have just repurchased and burned 1,772 TAO (~$445K) of SN4 alpha using organic revenues generated via Targon.com We have also burned an additional 6,678 TAO (~$1.7M) worth of SN4 alpha, which also represents revenue generated via Targon
42
99
70
494
70,898