Similarly, it blows my mind that all trees are made of air, specifically the carbon in it. I used to think that the biomass must come from the soil, but reality is more interesting.
It doesn't mention V shapes, though. I guess since a cyclist isn't generating lift, there are no wingtip vortices to exploit by forming such a shape.
> The shape or formation of the peloton changes according to multiple factors. Comparatively high power output efforts due to high-speeds on flat topography, a strong headwind or inclines (hills) tends to spread out or lengthen the formation, often into single file. A slow pace or brisk tailwind in which cyclists' power outputs are low result in compact formations such that riders ride side-by-side, often filling roads from one side to the other.
Ah, thanks for clarifying. Fenix 8 Solar is still MIPS but I can see why AMOLED is the "premium default" that they'd de-emphasize in their flagships. MIPS just shines less in the showroom, and doesn't have the phone display parity people expect.
I can't see them ever removing it from the Instinct line though, as that's the rugged one that signals tool.
Wait what? Do you have a source? I can't find anything about that, and I see the Instinct 3 is still being sold. Very disappointing if so, as that line has been the perfect pebble replacement for me.
Depends on personality I guess. That would be sooo unsatisfying to me. E.g. not wanting to accept that languages have exceptions "just because" is what got me interested in historical linguistics as a young lad.
At least for Nix itself, that's pretty much it except via Dutch.
> The name Nix is derived from the Dutch word niks, meaning nothing; build actions do not see anything that has not been explicitly declared as an input
It's the daydreaming/mind-wandering state that occurs when you're not focused on an external task. With all the stimuli of the modern world, I feel like we're being starved of crucial DMN time if we don't engineer conditions like the ones you describe.
It reminds me of how LLM hallucination is attributed to "I don't know" being underrepresented in training data, and it being a better strategy to guess on evaluations rather than admit not knowing.
Different reward function, but the same behaviour emerges.
No FFN is blowing my mind. This is pretty much "Attention Is ACTUALLY All You Need". Reminds me of BERT Q&A which would return indices into the input context, but even that had a FFN. Really exciting work.
Makes sense that the agent can refine its search terms/strategy based on discovered context.
But it still has to enumerate synonyms to find things.
I would assume it's very domain dependent, like code or technical docs would have more precise terminology that is better for fixed string search. On the other hand, medical or legal text can have many many ways to say something
Looks and feels premium, but ultimately fundamentally disposable.
This pattern extends to so many goods in modern life. Washing machines, microwaves, etc aren't worth the time of a local repairman. Repair is economically incompatible with its life cycle.
Clothes are replaced, not stitched. And after a few washes at that. Cars, phones, etc, consist of proprietary parts all sealed up.
Thanks for the perspective! I guess it depends on the outcomes in question
If they're measured by traditional academic metrics (parsing, recalling declension tables, translating into English), then Wheelock's grammar-first approach really does optimize for that. On the other hand Ørberg optimizes more for reading fluency and intuitive comprehension, which is harder to measure on a standard Latin exam.
If anyone is interested in learning it, there's nothing better than Ørberg's Lingua Latina per se Illustrata. It's entirely in Latin, including grammar explanations, but it starts out incredibly simple and ramps up gradually with lots of repetition. And that's fun AND effective, since you're immersed rather than grinding tables.