Are you sure you’re talking about fat and not total calories? The whole premise about Zone 2 running vs Zones 3/4 is you really do burn more grams of fat at the lower Zone per minute.
My house and street and surrounding streets still aren’t on Google Maps after a month post closing (new development). It’s on Apple Maps and Uber; Amazon and FedEx deliver.
Google Maps feels like I’m constantly in an A B test for how bad it can be and I’m always getting the bad side. It repeatedly doesn’t update in CarPlay, showing the same miles to destination as I progress.
It doesn’t label streets well either. I try to manually find routes in SF and I have to switch to Apple Maps because not enough streets are labeled no matter how much or little I zoom.
Ya and this leads to the point that if the simulation is so complex, which I take to mean is so like the real thing, then the near-perfect simulation and the actual thing must be highly similar physically.
And digital circuits are not physically similar to biological brains yet. So we shouldn’t conclude they have the similar consciousness property.
Well if you go that route, a computer simulating digestion has almost no physical features in common with actual digestion of a stomach. The same holds for consciousness and brains and computers. Them saying it’s just instructions is shorthand for pointing out the physical differences of brains and computers.
It’s all just particles, but the higher level differences are vast, and only brains are implicated for first person perspectives via science.
“The law of non contradiction exists”. Even Aristotle couldn’t “prove it” exists yet logic uses it all the time. I hardly think logic is about what exists but rather a tool, born out of interlocution.
Yes state schools are. That’s why it’s so competitive as well. $4-$6k at the state school vs $50K+ at private.
There’s one state school for my program within 100 miles of me (physical therapy assistant)
And a four year program is still 2 extra years of tuition even at the subsidized amount, and most would work fewer hours if at all because they need high GPAs.
To get into a radiation tech program, there are 260 applicants, almost all with all As, for 20 slots at my local community college.
Maybe in the very first instant you’d think it’s merit based. But, EVERYONE is playing the game. Getting homework and tests from friends who already took the class, taking classes at several different schools to get the easier teachers, paying multiple times the tuition cost on tutors and other study aides (eg $2k+ for all the anatomy models), every demographic is using paid ChatGPT. We all know which teachers to take. We spend much of class strategizing like this.
Every single student. It’s just another game to play or you lose.
1) True believers
2) Hype
3) A way to wash blatant copyright infringement
True believers are scary and can be taken advantage of. I played DOTA from 2005 on and beating pros is not enough for AGI belief. I get that the learning is more indirect than a deterministic decision tree, but the scaling limitations and gaps in types of knowledge that are ingestible makes AGI a pipe dream for my lifetime.
This is how I feel with AI math proofs. I’m not sure where they’re at now, but a year ago it took so much more time to check if an LLM proof was technically correct even if hard to understand, compared to a well structured human proof.
Maybe it was Timothy Gowers who commented on this.
Lots of human proofs have the unfortunate “creative leap” that isn’t fully explained but with some detectable subtlety. LLMs end up making large leaps too, but too often the subtle ways mathematicians think and communicate is lost, and so the proof becomes so much more laborious to check.
Like you don’t always see how a mathematician came up with some move or object to “try”, and to an LLM it appears random large creative leaps are the way to write proofs.
I guess I just don’t have the experience or optimism that a harness around an LLM, which can’t make the first, bare deduction on its own, is a good use of compute.
I got out of RLHF, including games and puzzles, before agents took off and maybe I have outdated info. But we estimated RLHF’ing a single hard full sized sudoku was ~25 hours worth of work.
https://cannonbol.com/blog/articles/Achten_%282004%29.pdf?ut...