rafaelero

8 ms·

91.
by rafaelero 3y ago
Local families are the ones getting rich by having their houses surge in price. And they are probably also the reason why new houses are not being built.
92.
by rafaelero 3y ago
What do you mean cities need it not to die? Would you say Monaco is dead? What about Zurich? Oslo?
93.
by rafaelero 3y ago
When those great role models are able to build a profitable spaceship company from the ground up I am sure we will pay attention to them.
94.
by rafaelero 3y ago
> I'll stick to the biologists, geneticists, and endocrinologists over a historian trying to sell books. You make it look like there's a consensus between them. It's unfortunate that Harari gets this bad press. I've r
95.
by rafaelero 3y ago
> Why would Harari's -- a historian's -- hot take be preferable to the consemsus ofof gender researchers and sociologists who actually study gender in experiments and statistical studies? Everyone can decide for themselves. Spe
96.
by rafaelero 3y ago
There are moments in which correlation rise, but over the long period the Pearson Coefficient is around 0.2.
97.
by rafaelero 3y ago
I don't think energy prices are important for BTC. If energy prices rise and it becomes unaffordable to mine, the network will adjust the difficulty level. The only input that matters for BTC performance is demand.
98.
by rafaelero 3y ago
It's his opinion based on his reading of the science, lol. And a reasonable one with which I mostly agree. If you think you have a better theory that explains gender differences, by all means, tell the world! But don't vilify peop
99.
by rafaelero 3y ago
Who would have guessed that an uncorrelated, scarce and highly liquid asset would leave asset managers salivating?
100.
by rafaelero 3y ago
And, curiously, the Alzheimer drug do work, just like hypothesized by the "dumb advocacy groups". You just have to take them earlier in the profession of the disease.
101.
by rafaelero 3y ago
That's what Phase II is for.
102.
by rafaelero 3y ago
The FDA could impose a ceiling over the price the company is allowed to charge if they are in a situation where there is reasonable expectation of efficacy but the trials haven't been consistent. Would that appease your concerns?
103.
by rafaelero 3y ago
I don't know and don't really care. It's about giving people the option, not forcing them to buy it.
104.
by rafaelero 3y ago
You don't have to worry about that, because companies are the ones financing it and if they don't believe the drug have a good chance of working they won't fund it.
105.
by rafaelero 3y ago
But what's the harm here? The drug is safe. The only burden patients had was financial. Now what if the drug was effective? The harm it would incur to patients by delaying it would be much more extreme.
106.
by rafaelero 3y ago
Nonsense. I'ld rather have a safe drug with >50% of chance of working approved than waiting for 3 more years and find out when I am already debilitated. That some drugs turn out to not work is not a reason to "lose trust in the
107.
by rafaelero 3y ago
MMLU is garbage. A lot of incorrect answers there.
108.
by rafaelero 3y ago
Bad articles like this is the reason why I keep buying Nvidia shares.
109.
by rafaelero 3y ago
They propagate information in a binary way (either they activate or not).
110.
by rafaelero 3y ago
https://deepmind.google/discover/blog/alphageometry-an-olymp...
111.
by rafaelero 3y ago
It's more likely that Nvidia will offer support to INT2 in the next generation and keep their dominance.
112.
by rafaelero 3y ago
LLM's can do math as well.
113.
by rafaelero 3y ago
Looks like we have finally rediscovered a biological neuron.
114.
by rafaelero 3y ago
Or you can just use a sunscreen everyday and get the best of both worlds :)
115.
by rafaelero 3y ago
You can train a smaller model and run inference multiple times and it will reach similar performance as a larger model running inference just once. What's the best way to make use of those multiple inferences is still up to debate, but
116.
by rafaelero 3y ago
You are incorrect. Increasing compute during inference renders similar gains to increasing parameters/compute during training time (see self-consistency, tree of thoughts, etc.)
117.
by rafaelero 3y ago
I remember thinking back then "well, how would I know better than the market?". Still, I did buy some NVIDIA stocks, which payed off great. My thought process comes down to: do I think people are as bullish as I think they should
118.
by rafaelero 3y ago
Even if 'regular' companies stopped demanding GPU's from NVIDIA, that demand would just transfer to cloud providers, which would keep buying those sweet H100. It's not a misguided AI strategy that is pumping the price. I
119.
by rafaelero 3y ago
Perplexity doesn't consistently decrease with context window size. Looks like Gemini reigns supreme for now.
120.
by rafaelero 3y ago
Mars?

More ›