Plenty fast for coding work and for sharing with my OpenClaw setup.

Currently in the process of adding another external GPU (RTX 4090 with pipeline parallelism) via thunderbolt 5 to the Olares One box, for higher quantization, possibly 8-bit, larger context, better concurrency, more kv cache.

It is still heavy as hell on my head, even with tons of 3rd party harnesses. On the other hand, my neck muscles are getting stronger.

Based on what I saw, and how visibly non-Chinese automakers are struggling to keep pace while looking over their rear view mirrors at the Chinese EV industry, my view is that the broader auto industry is heading for a major structural reset. Protectionism will slow it somewhat, but the broader unit economics will no longer support the current number of auto companies.

My prediction is that over the next 5 to 10 years, Mobility-as-a-Service, whether on-demand or subscription-based, will become the default model in dense urban markets. Private EV ownership will persist, but increasingly in suburban, rural, or less densely populated areas where shared mobility is less practical.

I have exited all my auto positions. And yes, Tesla should probably trade closer to a 30x P/E at most. A 300x-plus multiple is indeed disconnected from reality.

On the other hand, I'm keenly watching Waymo (and to a lesser extent Uber and Zoox), as potential future MaaS players perfect their autonomous vehicles/driver-less tech.

As he described it, people crammed into shophouses, kampongs (villages) and squatter settlements with no proper toilets (human faeces and urine were carted away by "night soil" men carrying them in open containers in the streets), no clean water, no drainage, no fire safety.

In 1959 barely 9% had public housing. The streets boiled over with riots, strikes and communist agitation, one bloody flashpoint after another.

Work was casual and wages were thin. The British still ruled but had lost all moral authority after the Japanese rolled over across the northern causeway with not much of a resistance from the brits (the idiots were stationed in the southern island of sentosa with their guns pointing south thinking the japanese will invade from the sea) and buggered them in the war.

Singapore was a poor, overcrowded, combustible place with no business surviving, let alone becoming a nation. The hard truth the world forgets: Singapore is an improbable nation. By all logic, it had no right to exist. No natural resources. No hinterland. No oil, no land, no army, no water of its own. Thrown out of Malaysia in 1965, a tiny island of immigrants with three races, four languages and nothing in the bank. By every textbook measure, it should have failed.

It didn't, because of one man's sheer will.

My father now is 90 years old, worked his way up as a menial laborer, put himself through night school, became a successful businessman, and built a family. To my father and his generation, LKY will always be their hero.

From a shit-hole to the first world. In one generation.

On the other hand, India’s progress continues to be hobbled by deep-rooted social challenges, including religious extremism, caste-based inequality, utter breakdown of civic culture, social fragmentation, linguistic chauvinism and regional rivalries.

Forced order or complete chaos. Choose your pick.

Vision may spark greatness, but execution is what makes it real. The Hacker News crowd can debate endlessly about who conceived the Macintosh while dumping on Steve Jobs.

But Steve Jobs did what ultimately mattered: he shipped.

Jony Ive without Steve Jobs micromanaging/editing/curating his design, is just another mediocre designer.

Had the privilege of meeting Jeff "Yak" Minter in Singapore, and also attended his presentation. Another legendary game developer, in the same league as David Theurer

The Zip drives that followed were abysmal. We sold a lot of them initially (I was working in a computer store in the early nineties), but sales cratered once the "click of death" became infamous. SyQuest drives suffered from similar reliability issues.

The founder of SyQuest, Syed Iftikhar, later left and set up another company called Castlewood, which introduced yet another removable drive called the Orb. The Orb was genuinely faster and more reliable than anything Iomega or SyQuest had offered.

But with the advent of cheap flash drives and faster broadband internet, transfer and storage of data shifted decisively away from spinning removable media, and the entire category quietly died, taking Castlewood, Iomega, Imation, and SyQuest with it.