🔗 Introducing BLOCKv V3 The next evolution of our Web3 Operating System is here, delivering enterprise-scale performance and zero-knowledge verifiability. Designed for the next wave of mass-scale, trustless Web3 applications. Learn more 👇 blockv.io/blog/blockv-v3…

This is going to unlock so much value and creativity. Ready Player One type vibes! Powered by $VEE 🌹🦋

the person who cares will almost always beat the person who's just doing their job

.London Stansted Airport You have a miniscule number of water taps to serve the ~80k passengers who pass through the airport each day. These taps also have a slow flow rate of less than a litre per minute. Apart from it being obvious, as an airport, you have whole teams who specialise in

GPT-5 was a cost play. Make a somewhat better model at a fraction of the price. ARC AGI shows it. GPT-5 is about 10% better than o3-pro but costs 90% less.

An extremely obvious version of AI + YouTube will look like personalized documentaries with video and AI agents linking the best most interesting clips Imagine a personalized docuvibe about Coltrane and Davis with specific video essay built just for the user on the spot

We need language to describe the "intent-agency-friendliness" of the web to recognise how many UIs are hostile to intentional attention. For example, on YouTube, if I don't want to see suggested videos related to prior curiousity, because this inhibits focus and flow states, I

this is the same story with ai. the killer apps will be the ones that collapse latency, friction, & context switching into something instant.

China is taking fireworks to a whole new level! Almost all the fireworks you see around the world, about 90 percent, are made in China!

An unforgettable moment. A volcano erupted just as a man was proposing to his girlfriend.

The real bottleneck in AI is office politics, not GPUs.

speed is overrated. clarity is underrated. you can sprint into a wall just as fast as you can sprint to pmf.

Golfer bags $2M+ thanks to a fly.

LLMs are non-deterministic, yet, demonstrate a kind of consistency in a model UNTIL said model is updated. Then results may differ. GPT 4o was amazing yet now it seems nerfed, alas. These optimisations of "using less compute" seem to mean reduced precision.

Apple's $1.1 trillion loss proves Steve Jobs was right about one thing: When you prioritize shareholders over products, you eventually lose both. Cook grew Apple market cap 10x, but he destroyed innovation. A company that once modernized industries now can't even finish Siri.

World Economic Forum's Yuval Noah Harari telling you what the so called "elite" have planned for the future and it's not good Don't say you weren't warned. They are very open with their plans.

Engineering graduates by country: China: 60k (PhD), 400k (MS), 1.5M (BS) India: 6k, 60k, 850k USA: 12k*, 50k*, 140k *50%+ of US MS/PhD are Chinese / Indian. Even if you assume the US ones are better, at +3SD, the best engineers in the world are overwhelmingly Chinese.

On the margin people are more worried than they should be about the “AI wrapper” question. The gap between a base model and integrating intro a critical business workflow tends to be pretty wide. Most enterprises will need highly tuned agents with a high degree of domain

Disruption never asks for permission.

With a few exceptions, it is harder to teach old dogs new tricks. I wrote this in 2010. It applies even more today with AI: "Young people, who are less entrenched in traditional ways of doing things, tend to adapt faster and figure out how to leverage new technology sooner than