gonedark

We've reached a major milestone in fully decentralized training: for the first time, we've demonstrated that a large language model can be split and trained across consumer devices connected over the internet - with no loss in speed or performance.

I was wondering myself for quite some time about what would be able to push BTC to new ath‘s, what would be able to top a bitcoin reserve or the president of the united states launching a memecoin. What if we are witnessing it exactly right now? If BTC really manages to stay kind of stable while SPX & Nasdaq are breaking down 11% in 2 days it shows exactly ONE thing: BTC needs to be a part of every diversified portfolio and isn‘t just some degen playtool - it deserves the attention of all kind of investors, no matter the risk appetite, age or anything else. More upside than most stocks, but also, at the same time, more stable in the toughest macro situations. This will result in A LOT of talks at A LOT of investmentfonds, millionaires and billionaires offices in the coming weeks & months. Cant wait to come back to this tweet in a few months. But first…everyone take care of yourselves. Dont look at the charts to much. Next monday has the potential to be another black monday. What if it would really only take a pretty major crisis to make BTC move from being a growth-like asset to a competitor to gold as store-of-value. Maybe that time is now.

We’ve shared a lot of updates today, but there’s one we know you’ve been waiting for. Here’s our leaderboard that measures the number of times we’ve said “AI” in the keynote. Looks like we’ve got a new front-runner. 😉

The Big Short against Retail (May, 2025) A few weeks ago, everyone called for a recession; now we just experienced nine consecutive green days in a row. I believe it's crucial to understand why this happened and why it won't last for long (nfa). Trump gave three main promises to the market: > Lower Rates (begging FED to cut last few days) > Lower Tariffs (mainly the deal with China) > Lower Taxes & these promises are expected to be kept, the market is pricing them in as a sure thing. Rates are expected to come down in July or earlier (Trump wants June), the deal with China is expected to be finalised before the containership recession trade comes into effect, and taxes are expected to be lowered with the successful tariff policy. These three promises (01) initially created this nine-day-long only up market. "who's laughing at retail now for buying the dip" ~citrini has been a quote that was stuck in my head for a while, so i decided to microanalyse each sector whether retail actually bought the dip, and, well yea most sectors are on a record high of retail activity which is honestly frightening. We're not only seeing a recovery since the tariff shock, but new ATHs close to everywhere. (02) And the data backs that. If we take a look at the 2-week rolling net flows of households we can clearly see that it's retail buying the dip (highest since 2019 pre-covid crash). (03) Retail investors have been turbo-conditioned to just buy the S&P 500 and win. DCA in the MSCI World and make 11% annually, and retire at 50. Thanks for playing. This conditioning has to be challenged at some point. Their belief system hasn't been challenged once; on the contrary, it has only been rewarded since COVID. The 4-year-long bull cycle turned into an infinite one that only experiences micro crashes and goes up every year forever. This will and has to stop at some point. As long as this conditioning of "buy every dip with size" exists, we won't actually see low prices, and we'll never know how bad it actually is until this buyer group eventually runs out of money, which they are right now by full-porting all their dry powder into QQQ. This is how we got to The large Meltup trade, a nine day long streak of consecutive green days on the S&P 500. (04) I argue that this hyper-conditioning to 'buy every dip' shows the market a narrative that is not true, which, as a consequence, changes how mechanical capital acts, creating a Thanksgiving turkey situation just waiting to unfold eventually (05 + 06). CTAs here are not per se "wrong", they just give retail investors the illusion of complacency, as they do right now with their $57b bid on top of retail accumulated shares. Eventually, all of this will unfold with the containership recession trade hitting the US in 5 days. All MAG7 earnings in the last earning season have been massively skewed and were useless information. TMT firms that manufacture physical hardware usually manufacture in waves, so the actual impacts will show up in their H2 capex over 1Q results. Thus, no tariff impact shown YET. I believe we're heading towards a bad summer, and hedge funds agree (07), am expecting a lower low on SPX eventually this summer, but a pristine Bitcoin. - EOY 2025 Targets - S&P500 - 4700 BTC - 110000 We're heading towards a sh*t macro situation. Hope you guys are prepared with cash to load up.

Absolutely convinced that decentralised training will be the single biggest narrative at the intersection of crypto and AI. That is the real decentralised AI tech.

Long long way to go for deAI clouds. Most problems that secured them a spot in the lowest category are fixable tho, so lets see where we are in a year from now. Oh and as expected, $GOOGL stonks go up 🙂‍↔️

TAO is fascinating in the same way that Ripple/XRP is. despite the cult energy behind it, absolutely no one (that I know of) has made a clear case for bittensor's tech stack without falling back on "incentives" as the core pillar. I would LOVE to engage with believers. dm me!

"will humanity ever do a 10 million GPU pre-training run?" OpenAI CEO, Sam Altman, raises the question: oAI employee: there'll be 10m GPUs working together on an AI system that learns and performs tasks. however, the approach may shift from fully synchronous pre-training to "semi-synchronous" or more decentralized methods.

I cant be the only one who is completely unimpressed by the agent (meme) meta, can i?! They are sooooo useless. The TAM is so small for them imo (at least in crypto). Majority of useful agents will be embedded into OS and @Apple is basically the only company able to win this race. Apple Intelligence is a start at least, the development within the next 5 years will be crazy to witness. I would bet money on Apple being the absolute agent-monopoly having iOS being utilized by 100s of millions of people on a daily basis. Crypto x agents is 99% pure fugazi, very very few expections.