O$IRI$
3.3K posts
O$IRI$
@cryptooosiris
Wannabe Trader Shitposter
Crypto's current state is a bit shit 1. Market cap is not an indicator of quality - the top 50 is made up of ghost coins or bloated governance slop that has underperformed and is uninvestable 2. The long tail speculative stuff went from high risk high reward to 'some dude in Miami is going to zero this if you hold it for more than 5.9 seconds' 3. Everything is extremely correlated and you can't meaningfully make bets based on sectors as it all converges into a tightly correlated mush, especially to the downside 4. Broad brush alt season is an artefact of the past that's very hard to replicate given (2) and given that there are simply too many coins and the excess of speculation doesn't really happen on centralised exchanges anymore - it's been siphoned off to bundled shit in max PvP settings 5. Crypto reputationally is no longer the sexy frontier of speculation. Institutional bid is in AI, retail speculative bid is in 0DTE equities, single name stocks etc. 6. Convexity has flattened. Even a lot of the historically safe blue chip stuff (BTC, ETH etc.) has underperformed and the historical anchor of 'buy deep drawdowns because all-time highs are guaranteed and explosive' has disappointed. All the shit we used to put up with because of the accessibly massive trend and momentum effects is now harder to justify because those same effects are getting neutered or siphoned off into other arenas. The obvious rebuttal is 'cycles' but even this past cycle is a useful counterpoint: it was extremely concentrated versus broad brush wealth effect, plus something very obviously broke after 10/10. So what does this all mean? 1. In previous cycles, nailing timing was enough and selection was the cherry on top (rising tide lifted all boats). I don't think that holds - both timing and selection matter now and in the future. 2. Participation alone can be an edge if the asset class is early enough and/or mispriced enough. I don't think that holds either, and we might actually have to learn how to trade (fuck). 3. Hopefully I'm an idiot doomposting the bottom GM
Most crypto projects have anonymous founders or ex-marketing guys. $TIG has something very different. Let me show you the team behind one of the most credible projects in crypto 🧵
Submissions to TIG have spiked in the last few weeks. The driver is obvious. AI is now strong enough to take a serious run at unsolved algorithmic problems, and TIG is the only place that pays for the result.
There's a quadrillion-dollar question at the heart of AI: Why are humans so much more sample efficient compared to LLM? There are three possible answers: 1. Architecture and hyperparameters (aka transformer vs whatever ‘algo’ cortical columns are implementing) 2. Learning rule (backprop vs whatever brain is doing) 3. Reward function @AdamMarblestone believes the answer is the reward function. ML likes to use pretty simple loss functions, like cross-entropy. These are easy to work with. But they might be too simple for sample-efficient learning. Adam thinks that, in humans, the large number of highly specialised cells in the ‘lizard brain’ might actually be encoding information for sophisticated loss functions, used for ‘training’ in the more sophisticated areas like the cortex and amygdala. Like: the human genome is barely 3 gigabytes (compare that to the TBs of parameters that encode frontier LLM weights). So how can it include all the information necessary to build highly intelligent learners? Well, if the key to sample-efficient learning resides in the loss function, even very complicated loss functions can still be expressed in a couple hundred lines of Python code.
Documentation on licensing soon @tigfoundation $TIG
BREAKING: Nvidia stock, $NVDA, officially hits a new record high, now worth $5.2 trillion. Nvidia has now added +$4.9 TRILLION in market cap since 2022.
One thing remains unproven for $TIG. The licensing model. Everything else is real: World class team ✅ Algorithm beats peer-reviewed 2025 paper ✅ Prometheus — millions of potential contributors ✅ One enterprise signs. The flywheel starts. $TIG
Prometheus. Knowhow stays on your machine.