I used to be able to curate my feed and pay attention to people I knew. I could see the photo projects my friends were working on, hear about life updates from past acquaintances, or reach out if someone was having a rough time.
Now, all cohorts, from recent coworkers to childhood friends I made before the first web browser, routinely spew paragraphs of LLM slop to shill a career coaching podcast. It’s Invasion of the Body Snatchers.
Social networks are only a small part of it. It’s email, mailing lists, billboards, sheets of paper stapled to utility poles, newspaper articles, dentist office phone lines, jigsaw puzzles, home furnishings, homework assigned to grade schoolers, birthday cards and on and on.
This is conflating two kinds of pink. The pink made from combining ends of the spectrum is most commonly termed ‘hot pink.’
The other, very often just ‘pink,’ is predominantly a light red. A quick and sloppy way to describe this is a light grey with a raised red component.
Also, you can make hot pink without needing to use spectral violet (the ‘end’ of the spectrum) since there are combinations of blue and red that are ‘metameric,’ creating a perceptually matching response in our eyes.
I'm not a hardware expert here but this strikes me as inaccurate, though the actual performance can be scenario dependent.
The Jetson hardware is targeted to low power robotics implementations.
The Jetson Orin is currently marketed as prototyping platform, and I believe it does not generally challenge recent Apple Silicon for inference performance, even considering prefill.
In the latest Blackwell based Jetson Thor, the key advantage over Apple Silicon is its capable FP4 tensor cores, which do indeed help with prefill. However, it also has half the memory bandwidth of an M4 Max, so this puts a big bottleneck on token generation with large context. If your use case did some kind of RAG lookup with very short responses then you might come out ahead using an optimized model, but for straightforward inference you are likely to lag behind Apple Silicon.
At this stage, professional inference solutions ideally use discrete GPUs that are far more capable than either, but those are a different class of monetary expense.
The story as told may be inaccurate but it wasn’t simply ‘what we had at the time’ either.
The 74 minute length resulted from Sony rejecting the Philips 60 minute 11.5cm diameter “Pinkeltje” disc size in favor of a 12cm diameter.
It’s quite possible that Sony’s Norio Ohga simply argued that the 9th symphony or various operas fitting would be enough of an advantage for the slight size increase without meaningfully decreasing portability.
I read Roger Penrose's The Emperor's New Mind when I was younger. It suggested quantum processes as a last ditch effort for a non-deterministic brain. At the time, I thought it was a fascinating prediction of how our minds might work and that reading it made me a smarter person.
I have since come to view it more as an interesting lesson in the pitfalls of hypothesis formation, popular non-fiction, and vanity.
Even so, as a layperson, it's entirely understandable to perk up whenever someone discovers 'tubules' in the brain, even if none of that sufficiently supports any of the collapse requirements of the Penrose/Hammeroff quantum microtubes.
He made hundreds of thousands of dollars from those streams, “twelve dollars” was ironic understatement about his dissatisfaction with the compensation Spotify pays artists.
This doesn't sound quite right, what's one of the names?
The London Dungeon is near the Eye and has been around since the 70s, albeit in different forms and location. They moved closer to the Eye due to rail station construction in 2013, I believe? Down the street from the old London Dungeons location is the London Bridge Experience, which generally gets better reviews and doesn't appear to be owned by Merlin Entertainment. I'd avoid both, but neither appear to match your description.