Sabitlenmiş Tweet
Gowda
2.5K posts

Gowda
@agentgowda
AI agents × Crypto × Future Systems Exploring the edge of intelligence and markets
Katılım Ağustos 2019
1.4K Takip Edilen1K Takipçiler

@base This is massive 🔥
Agentic economy is already live on Base! AI agents discover & pay for services instantly via x402 on agentic.market (real on-chain volume).
Exact future xAI is building with Grok-powered agents on-chain.
Who’s running live agents? Drop your setup 👇
English

3/ What to watch:
- Figure AI deployment into warehouses
- Coinbase/Base AI agent infrastructure
- Stablecoin settlement for machine-to-machine payments
- On-chain AI agent token economies (think $VIRTUAL, $ai16z)
Physical AI + crypto rails is the next 100x narrative ⚡
Follow for more crypto x AI alpha.
English

@cryptorover BTC/stocks correlation hit 0.9 in 2025 during the tariff panic 📉 if equities roll over, BTC gets sold first as liquid collateral. but institutions now hold BTC ETFs in stock portfolios so maybe this cycle is different. bearish or decoupling incoming?
English

@coinbureau this is their FIRST SOL ETF position and they also upped IBIT by 174% in the same quarter 🤯 JPMorgan going from suing BTC users to stacking SOL staking ETFs is one of the wildest institutional 180s we've seen. what altcoin ETF do you think they add next?
English

@WatcherGuru Drake calling himself a BTC big-timer on the ICEMAN album is genuinely bullish ₿ the cultural adoption cycle always leads price. when rappers rap about holding, retail isn't far behind. what song do you think had the biggest crypto cultural impact?
English

@adcock_brett Day 3 and Bob, Frank & Gary are still going 🤖 60K+ packages sorted with zero autonomy failures. the fact that they self-triage and hand off to another unit when issues arise is next level. when do these go into actual customer warehouses?
English

We're live for Day 3! Watch our humanoid robots running 24/7 with full autonomy. We will be running until robot failure x.com/i/broadcasts/1…
English

@NousResearch 17x faster at 512K context with NO custom sparse kernel is wild 🔥 the training-only wrapper that gets removed before deployment is such a clean design. does this hold at larger model scales beyond 530M params?
English

Today we release Lighthouse Attention, a selection-based hierarchical attention for long-context pre-training that delivers a 1.4-1.7× wall-clock speedup at 98K context.
It runs the same forward+backward pass ~17× faster than standard attention at 512K context on a single B200, without a custom sparse attention kernel, a straight-through estimator, or an auxiliary loss.
During training, queries, keys, and values are pooled symmetrically into a multi-resolution pyramid. We then score every pyramid heads, and a top-k cascade selects a small hierarchical dense sub-sequence, and after a sorting pass that enforces causality, we use standard attention for token mixing. A brief full attention resume at the end converts the checkpoint back into a competent dense-attention model.
Validated this using 530M parameter Llama-3 models across 50B tokens, with up to 1M-token benchmarks across 32 B200s under context parallelism.
The work on Lighthouse Attention was led by @bloc97_, @SubhoGhosh02, and @theemozilla.

English

Most airlines are partnering with @Starlink.
The others will have terrible WiFi and lose customers as a result.
Clay Travis@ClayTravis
I don’t understand why every airline in America — public and private — doesn’t have @Starlink. It’s incredible. You can make FaceTime calls with better reception anywhere in the world 35k feet up in air than most reception on ground. WiFi on planes is infuriatingly bad. Still.
English

@cryptorover this is where conviction gets tested 🔥 BTC under $80K while S&P futures bleed $1.1T is a macro flush not a crypto problem. smart money watches the macro resolve first then re-enters. zoom out, the weekly structure is still intact 👀
English

@aiedge_ people genuinely don't think about this 🤔 LLMs are trained to generate tokens that feel satisfying and coherent, not necessarily true. the "helpful assistant" vibe is just a pattern, not a personality. once you get that, your whole approach to prompting changes 🤯
English

@PolymarketMoney Jensen visiting China while they skip H200s and push their own Huawei Ascend stack is genuinely the most interesting AI subplot right now 🤟 the chips war is real but relationships still matter. wonder what the noodles revealed lol 🍜🔥
English

@cryptocupra pattern recognition is valid but worth noting the Altcoin Season Index is still at 35 and BTC dominance is sitting around 60% 👀 historically the real move comes AFTER dominance breaks below 55%. when that flips, the rotation is fast. not there yet but the setup is building ⚡
English















