Hmm I wonder what the preselection to get to the top 10 are? Human readers or automated?
190 karma · joined July 8, 2017
It'll happen a week or a month after IPO date though? It took fb/meta 1 year and then it entered as 1% qqq. TSLA entered 3 years after IPO so probably a small percentage.
Tsla is 2% vti (2T AUM). QQQ is 400B AUM. So add those two and you get $56B of purchasing. This seems like the amount they want to raise via IPO in total in the news, so the banks who do the IPO can sell it all guaranteed.
But people will want to buy it before it gets into the passive funds... So... Post inclusion market cap will be higher than we expect?
To me it seems the opposite: Rent control means supply goes down, so available building & land prices go up. These prices going up means an opportunity for builders who are good businessmen because they are going to make a margin on their investment, the bigger the investment the bigger upside.
Another intuition is with rent control it's hard to extract new value from an old building, so that also incentivizes tearing it down and squeezing more units into the land.
In SF, rent control exists on all buildings built before 1979. It appears to me that people who prioritize new builds pay a huge premium for them. I think this particular rule also incentivizes tearing pre 1979 structures down, vs the no rent control newer buildings can continue to have growth in the value extracted from them.
Anything wrong with that?
I've been using it for about a year.
My impression is if you fly for work, you get a lot of employer sponsored points, so it's interesting.
But if I fly 5-10 trips a year personally, why would I try points when I can get 3-5% cash back on my various cards?
I had a HW assignment in 2017 to do this, specifically we were given raw wifi signals and told to determine how many people were in the room. We did it by detecting breath rate of each person with the MUSIC spectral algorithm. For that challenge there ended up being 6 people in the room, all with slightly different breath rate (attenuating the signal as their chest expanded).
Is it because you don't need to buy many gpus to do your workload?
Some are UI tests, but we don't recommend those, because we found they don't catch breakages as often so we don't support the language they're written in. The tests we recommend are often integration type tests in that they call very higher order functions and often many of them.
We do it on a symbol level after statically analyzing each change, and everything in the monorepo daily. Our remedy to high risk changes is to run more tests, client tests not unit tests. Sometimes there are 100k client tests to pick, so we rank them and run a small subset.
It is a hard problem. One interesting observation is that there is a culprit symbol or two in the culprit change, but its connectivity is very similar to non culprits in the same change.
Another observation is that the transitively modified callgraph after a change is pretty big, a depth of 50 is not unusual. It is hard to get many useful signals out of it beyond amount of overlap in transitively affected symbols between change and test.
We found file level and build target level to be too coarse, but AST symbols are working.
What would be the difference in user experience or information retrieval performance between the two?
My impression is it saves work on the dataset transformation and compute for fine tuning, so it must be less performant. Is there a reason to prefer the strategy here other than ease of setup?
Or GitHub copilot?
I had a Prof say "no matter how I clown around up here, you're going to have to teach yourself these things".
My perspective is that classes provide the motivation, then each student does what they do with it.
I don't think departments are financially incentivized to create happy customers, and I do still think this is a good thing after my own multi-year "sufferings."
Molehill: reciprocal of all points in a circle on the complex plane.