Manfredas (@manfredasmi) header
Manfredas (@manfredasmi) profile picture

🍄

Joined October 2014
Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)Photo by Manfredas (@manfredasmi)
Been building a personal health tracking app for 3 months mostly for myself: - import blood test PDFs - track trends over time - cross-reference with workouts and supplements - ask Claude AI questions about your own data If anyone wants to try it: myapex.vercel.app/🙏
1
10
102
Update: out-of-range markers on my own bloods went from 13 (May) to 3 (Oct) using Apex to track trends. Super happy with the results 🙌
1
Manfredas retweeted
POV: you’re a ethical person in crypto
100
48
7
649
26,122
One of the main reasons I decided to hold zero-gravity:native in the AI infrastructure sector is that 0G Labs launched $iAI. enice's $VVV = 0G's zero-gravity:native = AI compute finance model Venice ($VVV): $1.75B 0G (zero-gravity:native): 243.4M Now, let's compare their AI compute tokens through staking 1. $DIEM (Venice) = $1667, which gives $1 daily credit Annual usage value: $365 2. $IAI (0G) = $390, which gives $1.271 daily credits Annual usage value > $450 So basically, $IAI is currently trading at a 76.6% lower price than $DIEM, despite offering a higher daily AI usage value That's why I believe demand for iAI will grow, and it's one of the main reasons I invested in zero-gravity:native You can mint iAI using 0G as collateral so I expect a huge portion of the zero-gravity:native supply to be staked $VVV already managed to gain market attention and a much higher valuation through the AI compute finance model along with buyback & burns Meanwhile, 0G is still in the early stages of expanding its iAI ecosystem Bear markets are for finding hidden gems
Alpha I heard at TOKEN2049: The poor panic sell. The rich accumulate. If you don't hold privacy tokens, you'll be poor. If you don't hold AI infrastructure tokens, you'll be poor. If you don't hold DEX tokens, you'll be poor.
65
19
3
159
8,290
At launch: lending via Aave/Fluid, liquid staking/restaking via Lido/EtherFi, and vaults across Morpho, Euler, Gearbox, Spark Execution via one-click "Zaps" swap + bridge + deposit bundled into a single tx
1
31
Not yet live: Pendle PT positions (~2 months post-launch per roadmap) and perps LP tokens Both flagged as upcoming, not shipped.
11
Manfredas retweeted
I am done with this shit. It is over. The state of engineering right now is horrible. It has been half a month since I started a new role at a big company. Nobody knows anything here. The specs, code, tests, PRDs, tickets, resolution of those tickets, reports, etc., everything is made by Claude Code. Nobody on my team likes this. They are being forced to ship as much as they can. I have heard multiple times from higher management that pushing code is not a bottleneck, so why are we slow? People are working 12 to 13 hours a day just to press enter. Nobody is reading anything. Humans in corporate are doing nothing on their own. Everyone, literally everyone, from an L1 to an L7 engineer here is doing the same thing. Talk to Claude. There is no sense of victory. Nobody is resolving bugs. In reality, nobody is thinking anymore. Everything is done by LLMs. It is so soul-sucking. I would not mind it, to be honest, if we were at least given the time to check out the code and see what is going where. But no, the goal is to just ship. No matter what happens.
2,029
4,736
1,669
48,573
8,435,746
Eisman is spot on. The frontier AI labs are manufacturing a crisis with the goal of regulatory capture.
31
92
7
396
36,251
I'm usually not one to write thought pieces without much technical depth. But here we go. Slow the fuck down. mariozechner.at/posts/2026-0…
150
491
176
2,864
630,091
Manfredas retweeted
You all know by now I’m a big fan of $TAO, and subnets in particular, so let me share something. At this point, there is one subnet that really stands out to me. That subnet is Subnet 44 - Score. I’ve reiterated several times that I believe this will be the first breakout subnet. Let me explain why. Most Bittensor subnets operate in markets where margins are tough. Even if you look at projects like Chutes, which I genuinely think is a strong project, selling compute with meaningful margins is hard. At the end of the day, compute is still compute. Even with a solid integration layer on top, you’re largely reselling infrastructure, and margins remain thin. With Score, that dynamic is different. Their edge is the product itself. They are serving an industry that, for the most part, doesn’t fully exist yet. Yes, some companies use certain forms of Vision AI, but many still rely heavily on human annotators, or more importantly, they don’t use their camera footage at all. Take a supermarket chain as an example. They already have cameras installed everywhere. But those cameras are mostly used reactively, to review incidents after something goes wrong. The untapped potential is enormous. Just off the top of my head: 🔸 Fully automated inventory resupply 🔸 Customer behavior analytics (where do customers stop, what do they look at, for how long, and what they ultimately choose) 🔸 Staffing optimisation 🔸 Automated shoplifting detection And these are just surface-level examples, I’m not even a specialist in this field. The core issue today is cost. Using camera footage intelligently is simply too expensive for most companies. Thats where Score changes the equation. By making Vision AI better, faster and cheaper, they unlock entirely new value streams. There is a massive amount of underutilized camera data globally and Score is positioning itself to monetize that inefficiency. What gives me real conviction is this: They already have paying customers, which they are already delivering real world value. And they are doing this before Manako is even live, which will directly, and fully automated by code in the form of a manifest, translate real world partnership tasks to miner challenges. This will significantly accelerate their whole operation. To go back to my earlier point: Score isn’t just aiming to break even against miner emissions. They are building with margin in mind. They are building to be profitable. At the moment, no other subnet is operating at this level. Sure, some are progressing well, but moving from a decent product to scalable, paying enterprise customers is a massive leap. Score is miles ahead, both on landing enterprise customers and on becoming actually profitable.
15
33
4
199
12,149
Manfredas retweeted
Your AI calls go through someone else's servers. They log your prompts. They train on your data. You agreed to it in a ToS you never read Today I routed my OpenClaw agent fleet through @0G_labs compute running GLM-5 wallet-authenticated, no middleman, no prompt logging. One of the best open models, fully private.
6
4
25
6,293
Crypto as we know it is over. It's time to look at the abyss and stop acting like this industry will in any way recover if the leaders continue to act like everything is fine and the 50th perps DEX will solve this. The Death of the Crypto VC sector is slowly unfolding during the past few months. LP commitments have been on a low and didn't even remotely recover during a generational $16k -> $120k bull market. VCs like Mechanism/Tangent literally pivoted away from crypto. Half of the Paradigm team ragequit in the last 2 months, entire firms silently exited everything. Barely any crypto VC has been able to raise for another fund and the venture appetite is close to zero. Please for the love of god look at the data & stop coping that this in any way is normal or will recover for a sector claiming to be on the frontier of technology. The risk appetite inside venture has been off the charts in the last 3 years, blockchain received only outflows. I spoke to so many VCs (both tradfi and crypto) in the past month, and close to nobody was optimistic about them being able to raise for another crypto-fund. We are at the tail end of blockchain innovation. "Oh ownership coins fix this" No they don't. Sorry to burst your bubble, but as the founder of a company doing "ownership" structures, this fixes exactly nothing. It's a band-aid of complacency. I'd argue it actually makes it worse, because no talented young founder will chose to give anonymous tokenholders full control of their business, it just turns crypto even more into this autistic cypherpunk delusion. Blockchain & especially alt coins has moved from the frontier of technology to an un-investable asset class who's building products who nobody needs. And the VCs who are left are trying their best to unauthentically manufacture narratives, fund the current hot thing (just to be left at 0 after the 3y vesting starts, and the current hot thing turned out to be not so societally important as the fast moving crypto sector thought it would be). The frontier of technology has moved away from blockchain and sits at AI & Robotics right now and blockchain right now is seen as the weird industry you enter to build something meaningless for exit liquidity. If we want this industry to bloom again, we need to work to get rid of the 3 in web3 and come back to reality. We need to go towards the epicenter of the current innovation and not try to artificially replicate it inside crypto. It's either valuable tokens for web2 startups or this sector & especially the venture market goes to 0. @StreetFDN
[ ZOOMER ] KYLE SAMANI STEPS DOWN FROM MULTICOIN, MOVES ONTO OTHER TECH AREAS
243
142
103
1,583
409,140
Manfredas retweeted
Two milestones. One weekend. - $2M all-time volume - $1M monthly volume (first ever) The momentum is undeniable: 48% of all lines placed in 26 days. A full quarter of growth compressed into one month. This execution frequency & volume is systematically filtering out 'luck' from the outcomes, turning theoretical edge into realised returns. January's ~6% ROI is the early signal: more data points → clearer edge realisation. We are seeing the results of a machine operating exactly as designed. Accelerate.
10
14
1
71
3,993
Manfredas retweeted
αVault: Monthly Volume Reaches All-Time Highs January has already reached an ATH in both monthly volume and execution lines, with over a week of trading still to go. This momentum is a direct result of the model's recent strategic refinement, which has increased our capacity to identify and capture market inefficiencies. Since our first full month of trading in November, αVault has seen a 30% increase in average daily volume (normalised to a consistent bankroll for comparison): - Nov: ~$30,000 / day - Jan: ~$39,000 / day (ATH) As we outlined in a recent post, part of the model's refinement is to scale execution through line diversification, increasing the number of statistically independent +EV positions rather than sizing up individual bets. By doing so, we spread risk across an increasingly diversified portfolio. This systematic scaling is already netting results: we have gained +$37k in PnL so far in January, recovering the bulk of the drawdown from late last year and validating the framework’s performance at scale. Volume at ATH. Lines at ATH. PnL recovering. We are just starting to realise the full deployment capacity of this framework.
1
9
3
54
2,212
Manfredas retweeted
Score x @ReadingFC Over the next year with @ReadingFC, we’ll build vision AI models that turn match footage into game changing insights for their coaches and players, and systems that make stadium operations smarter and more efficient for elite sport organisations.
57
77
61
448
437,999