Nikifers

The best project I've found so far this cycle : TensorCash, the verification and settlement layer for global AI inference. The game of ever evolving liquidity has meant 99% of traders are absolutely rinsed chasing FOMO KOLs , bundled claude coded slop and new narratives daily. Nowadays, there is barely any plays that go up more than 24 hours because every noob in here turned into a deployer and there's more devs than traders. The game is solved, your make it days are essentially over. Most of you guys will NEVER even make a cent from this game, its obvious . Unless you're a top 0.01% trencher simply put u are not going to make it. Even if you deploy ain't nobody going to buy your little coins, back to McDonalds applications unless... If you're like me and not interested in this dumb ass fortnite bundle PvP between 100 chains on 100 launchpads this may be the play for you. If you're tired of bundled memes, larp tech and launchpads on 123 chains and want something interesting to research, see tensorcash:native tensorcash. =========== What is tensorcash:native and why are so bullish mr popseye? ===== Simply put, I look at the narratives in real life. Not some dumb Elon tweet, a zoo meme or whatever silly agent narrative ur KOL came up with on the spot to farm you. Rn its obvious the world's taken over by AI, AI AI, it's all you hear. The big part of AI that most people have to come to realise that’s important is inference. Recently in crypto inference marketplaces have gone to 100m and projects that actually have real world implications built on blockchain are going to 5b + rn. POD is sitting at 300m FDV , VVV is at billions. Anything inference related deserves to have attention. The market and rightfully so is hungry for inference projects and legitimate ones at that. In any work we do nowadays, we use AI, AI models, LLMs, everything NEEDS and will continue to need AI. That part is undebatable. However, the important part of these AI models is the inference that it uses. But as more and more providers, customers, teams, companies use these AI models, how can they be sure the inference they pay for is what they are getting. Eg claude has several different models with diff amount of tokens, how smart the model is (fable vs haiku). Any provider can just charge u for an expensive model and give u a cheap one (I don’t mean claude but its an example of how differing models cost different). Actually its proven this is the case for almost half the time you pay for a model, you are being lied to and sold a cheaper one. The AI model and inference LIE Researchers auditing third-party API resellers found almost half of their fingerprint tests (45.83%) failed to match the advertised mode. Source: arxiv.org/abs/2603.01919 . TSC is a layer 1 blockchain that ensures these models are actually what you are paying for. It makes the underlying computation independently checkable. The computation can also serve a paying customer, allowing miners to earn service fees AND mining rewards from the same job. It's a permissionless L1 blockchain where miners, instead of wasting resources in brute forcing the next block, are running trained AI models, a process called inference, and compete for TSC That gives AI providers more room to compete on prices, while still covering their hardware costs. For the buyers, verification gives a reason to consider an unknown provider: there is evidence of what ran, beyond the company’s name on the bill. It's a form of decentralization of inference, because you don't have to trust anymore, you can verify. (quoted from anapneo) In summary - Providers (datacentres, GPU operators, rotating BTC miners) sell AI inference - Users (apps, agents, businesses) buy it via AskTilly and other marketplaces - TSC sits underneath as the trust layer, proving the AI work you paid for was actually performed - Clients pay fiat → AskTilly and other marketplaces buys TSC and pays miners → miners sell less → price up → mining more profitable → more miners, bigger network → more clients Plenty of Web3 infrastructure projects build impressive technology but ultimately remain stuck inside Web3, waiting for someone to actually use it. TensorCash already has AskTilly being built around its verification infrastructure. AskTilly is being built by Enrico Ottavian, who spent 11+ years at Goldman Sachs and became a Global Head in macro/rates structuring before moving into AI markets. Again , already there has been INSTUTIONAL players using tensorcash. Asktilly, by Enrico Ottavian ex Goldman Sachs and Samad Javadi ex morgan Stanley, efinancialcareers.jp/news/th… is part the ecosystem with the network produces over $12M a month of inference at market prices, . boblabs.eu/runs a pool. It is of my opinion this continues to get more and more adoption from instituions. Not crypto native audiences because this is a project for real world catalysts not the poop broke world of web3. On the technical side, Ismael Aadnan / Bob Labs is working directly around the TensorCash ecosystem and inference infrastructure. So you've got a fairly unusual mix: AI engineers +PoW/crypto infrastructure + Goldman structuring experience + Morgan Stanley+ institutional capital networks There have also been conversations with some much bigger names Again the target isn't crypto native audiences because this is a project for real world catalysts not the small broke world of web3 where we sit there waiting for the dev from the jungles of Vietnam or India to not rug us for 300 bucks. Putting it simply: > Provider: “I ran the model you paid for.” > TensorCash: “Here's the proof.” Personally at 20m circ and with very low emissions I look at other similar projects going to billions (and probably going higher) and I think this is an insane opportunity and actual tech to get behind for once. Prl rightfully so is getting big attention at token2049, so it looks like inference projects are getting buzz. Criniti just posted an article about AI x crypto. Again the target is tradfi and instituions. This is solving a projects 300B+ dollar industry going into probably trillions soon at this rate with institutional clients already and continuing to show interest and similar projects already at billions. Imagine the blockchain that solves the biggest pain point for AI and customers and providers atm. Remember olas at the start u had no one talking about it except two random accounts with 0 followers. Well similar here. There's definitely been some groups that have talked about it but realistically on X only one other person has found it and made a good thread. (borrowed a few quotes as he is so good at writing). nitter.cf/anapneo/status/2106376… If you are interested in buying its only available on safe trade atm safetrade.com/exchange/TSC-U… I would also recommend using LIMIT orders, do not market buy as the orderbooks are thin and it takes time to accumulate. I would wait for a dip using limit orders. Please do not market buy click as the orderbooks are thin and get slipped hard. There are other things about this but I need to write another thread as this got too long nitter.cf/Tensorcash