I’ve been something of a gym rat since I was a teenager but sometimes the best way to get your heart pumping and your brain thinking is to get out of the gym and onto a hiking trail (1/10) I love going for long hikes that get your blood flowing. I travel a lot and wherever possible, my assistants identify local opportunities for some good outdoor treks (just search my X account for ‘hike’ and you’ll see what I mean) (2/10)

High-performance computing (HPC) sites are looking to nuclear power to fuel their electricity demands, which I’m all in favour of, provided there’s a wider societal benefit from these HPC operations. Otherwise, shut ’em down (1/13) Tech giants like Amazon, Google and Microsoft are making deals to ensure access to power generated by nuclear power plants—big facilities like Three Mile Island but also small modular reactors (SMRs)—to keep their AI data centers humming (2/13)

GREAT NEWS – the @BSV_Assn launched the World Wide Web Consortium (W3C) Blockchain2 Community Group to explore the integration of enterprise blockchain technology into internet standards for web apps (1/12)

bsvblockchain.org/news/blockchai… The W3C was founded in 1994 with the goal of developing standards for the ever-evolving web. That long predates Satoshi Nakamoto’s introduction of blockchain tech in 2008, but the two technologies are now firmly on a collision course (2/12)

Somebody sent me a Darrell Green highlight video the other day that showed the ‘ageless wonder’ running down guys nearly half its age. Got me thinking about the need to show up for work every single day (1/9)
For those not familiar with North American football, Green played cornerback (CB), a position requiring blazing speed, sharp eyes and quick decisions. It’s rare for a CB to be still starting past age 30, yet Green was still a starter when he was 39 and didn’t retire until 42 (2/9)

The failure to introduce AI-specific regulatory guardrails will almost certainly come at a cost to the public, much like the failure to rein in the worst impulses of ‘crypto’ operators (1/13) California Gov. Gavin Newsom recently vetoed legislation that would have imposed (in my opinion) modest oversight of the state’s AI operators, handing a victory to the Silicon Valley giants that lobbied heavily against the bill’s passage (2/13)
gov.ca.gov/wp-content/upl…

Baby hippos are eating ‘crypto’ and it won’t be pretty when the utility-free memecoins get fired out the other end (1/14) MOODENG, a memecoin named after a viral baby pygmy hippo, saw its market cap surge past US$320 million just one week after its launch. And because imitation is the sincerest form of grifting, copycat hippo-themed tokens quickly followed in MOODENG’s wake (2/14)

Mark Zuckerberg says content creators are overrated in terms of their data being scraped by training AI models, but the day is coming when his Meta AI model will have to pay its way (1/13) Zuckerberg believes “individual creators or publishers tend to overestimate the value of their specific content,” and while “certain partnerships” will be made “when content is really important and valuable,” Zuck says he’s not bothered if individuals want to opt out (2/13)
theverge.com/2024/9/25/2425…

When I was a kid, my father told me that the worst day of being your own boss was better than the best day of working for someone else. It was the best piece of advice I ever got (1/15) When my dad married my mom, he borrowed some money from his new father-in-law to buy a small farm near Lashburn, Saskatchewan. He got into the registered seed business, produced a bumper crop and paid off the loan in no time. From that point on, my dad was his own man (2/15)

The concept of ‘truth’ will become ever more elusive as inaccurate AI-generated content floods the web, creating authoritative-sounding but incorrect info to be ingested by other AI large language models (LLMs), making each iteration more corrupted than the last (1/12) A growing body of research shows that LLMs fed by AI-generated data leads to increasingly nonsensical output, aka ‘model collapse,’ in which AI cannibalizes itself until accuracy has officially left the building (2/12)
nature.com/articles/d4158…

A recent thread on this platform tries to explain why so many ‘crypto’ boosters are far less engaged than they were in the past regarding the ability of bandwidth-constrained blockchains to solve any real-world problems 1/11

https://x.com/Travis_Kling/status/1830981118566244633

The thread is #3 in a series, following previous efforts noting the growing ‘Lack of Pretense That Any of This Shit Does Anything or Will Ever Do Anything’ and how ‘Financial Nihilism’ drives younger demographics into speculative crypto bets 2/11

The annual Ig Nobel Prizes—issued to scientific research that offers dubious value to society—has once again missed the boat by declining to celebrate studies that examine why ‘crypto’ prices go up or down 1/14
arstechnica.com/science/2024/0… The forces behind up or down token trajectories remain a mystery to most token holders, so much so that crypto news sites repeatedly run articles attempting to explain why certain tokens are up/down today (before ultimately declaring ‘we have no clue’) 2/14

I’m fortunate enough to own properties in several countries, some of which I bought already built and modified to my tastes, others were built from scratch to my specifications. I’m no architect, but I’ve always found property design fascinating and a lot of fun (1/9) I’m a generalist so I’m not bad at a number of things, but I can only take an idea to a certain point before I need to hand it off to someone who specializes in whatever area we’re talking about and let them carry the ball the rest of the way (2/9)

More and more BTC block reward miners are ‘pivoting to AI.’ But it’s not because AI is the new hot chick at the tech sector bar. It’s because the block reward economic model is broken (1/10) BTC’s failure to scale makes it impossible for the network to realize Satoshi Nakamoto’s plan for transaction fees to take the place of ever-diminishing block rewards as miners’ primary source of revenue. (2/10)