surrealize

Of course increasing supply lowers the equilibrium price, relative to the case where supply doesn't increase.

https://research.upjohn.org/cgi/viewcontent.cgi?article=1325...

> This suggests that new construction reduces demand and loosens the housing market in lowand middle-income areas, even in the short run.

All of us are making assumptions about how much testing Tesla does internally before deploying a feature to customer cars. I think it's way more than the equivalent of "3 test patients", but from the outside of course there's no way to know.

From the linked tweet, there's no way to evaluate how unusual the failure case was. And of course we don't know how likely it was to have been a pedal misapplication. We do know that that's not uncommon, though.

The original comment I replied to assumed that there was no internal testing, and you're assuming there was very little internal testing. Both of you are jumping to that conclusion with very little support.

Of course Tesla tests things itself with its own drivers. What they can't do is replicate every single one of the unusual corner cases that people encounter in the real world.

Right after someone dies, there may be people mourning them, so it's terribly disrespectful to criticize the dead person. That concern about disrespect is largely about respecting the mourning process.

But, after some time has passed, we should be able to discuss and evaluate a person's impact (good and bad). Otherwise history would be impossible to write.

It's been eight months. I think it's okay to acknowledge the full range of what Davis said.

Grandparent comment didn't say Davis was terrible, as you claim; the words were "He said some terrible stuff".

> But beyond that, I think even with Rust's macro system it could be difficult to make it work for Tensorflow in a way that feels appropriate for Rust programmers _and_ for TensorFlow.

If you're talking about matrix shape compatibility (matching up rows from one with columns from another) I'm hopeful about const generics here: https://github.com/rust-lang/rfcs/blob/master/text/2000-cons...

Forcing open APIs isn't anti-business. It's pro-competition and pro-startup. The only thing it's anti- is big incumbents, which is the point.

As more agents enter the business, a lot of their time and effort ends up going into getting clients in the first place. You end up with lots of agents spending more of their time trying to get business, with actual home-sales work taking up a smaller proportion of their time. But the increase in home prices is enough to make up for the decrease in the number of homes sold per agent.

Point being, an increase in home prices, and the resulting increase in commissions (in dollar terms) doesn't necessarily result in any benefit to clients.

In the system you envision without stacking, are all of the large blocks suspended from the crane simultaneously when the system is fully "charged"? That would require the crane to support a much higher max load.

> so long as investors have been alerted about which social media will be used to disseminate such information.

They have been:

> Item 8.01. Other Events.

> Tesla Disclosure Channels To Disseminate Information

...

> For additional information, please follow Elon Musk’s and Tesla’s Twitter accounts: twitter.com/elonmusk and twitter.com/TeslaMotors

https://www.sec.gov/Archives/edgar/data/1318605/000119312513...

from the BBC piece here: https://www.bbc.com/news/business-45105674

I don't know if it'll happen, but it's been proposed, e.g. https://www.nbcnews.com/business/energy/china-unveils-propos...

I saw how badly he treated people in the rust community, so I have zero doubt that his problems dealing with people played a role in the current situation.

Didn't intend to belittle his contributions though; he contributed significantly to rust and (as I understand it) did ~everything technically on CopperheadOS.

It's another example of Daniel Micay (strcat) contributing technically to something, having a falling out with other folks involved with it, then quitting in a huff.

Not a huge surprise if you followed rust a few years back:

http://slash-r-slash-rust.github.io/archived/2u1dme.html

I, for one, welcome our new inside-joke-forgetting overlords

Local control (and the perverse incentives that operate at that level) is exactly what's causing the problem. State-level intervention is the achievable way to fix it.

Homeowners in a city have an incentive to block housing construction. That hurts everyone who commutes into that city; those people deserve a say too, and they have one at the state level.

It's not a significant direct cost. Millions of extra DNS queries amounts to single-digit dollars (route 53 pricing).

It's an annualized rate of hundreds of thousands if you include all the models. For the model 3, 2600/wk makes for well over 100k/yr. Model S and X combined are going at an annualized rate of about 100k/yr, extrapolating from last quarter's 24,728 number. Altogether it's over 200k annualized.

I thought Bojack Horseman was pretty original.

So you have maps that record the location of stationary signs etc. and then have the car stop if it gets a radar return from a stationary object (e.g. fire truck) that's not on your map? Then when someone puts up a new sign you're slamming on the brakes until the map gets updated.

Not being able to tell if a stationary object is something you're going to hit or not (like LiDAR or stereo cameras might do) is more relevant to this case than maps.

You mixed up some units here: Brooklyn is way denser than SF (about twice as dense). Your SF number is people per square mile, not per square km.

Same units, Brooklyn is ~35,300/m^2, SF ~18,500/m^2

Manhattan ~69,400/m^2

all from https://en.wikipedia.org/wiki/List_of_United_States_cities_b...

What is up with the NYT being negative about the tech sector? Especially on the west coast. Can you imagine the headline:

"Manhattan’s Skyline, Now Inescapably Transformed by Finance"

Presumably you mean they could follow one another very closely. If they're traveling at several hundred mph, that doesn't give following vehicles much time to react if there's a problem with a leading vehicle.

> I don't think air resistance is the largest cause of expense in transportation.

If you want to go fast (the plan for the hyperloop) then air resistance becomes a significant cost. The power it takes to overcome aerodynamic drag increases with the cube of velocity.

I imagine the hyperloop design where cars are independent makes drag costs worse compared to trains (compared at the same speed).

I'm generally against concentrated market power, so I'm definitely sympathetic to the point and I think it's good to be aware of this concentration of ownership.

I'm not sure this is totally fair, though:

> We found that the Big Three, taken together, have become the largest shareholder in 40% of all publicly listed firms in the United States.

> Together, the Big Three are the largest single shareholder in almost 90% of S&P 500 firms

> The Big Three – seen together – are virtually always the largest shareholder in the few competitors that remain in these sectors.

They're aggregating these owners together, then saying that the aggregation is larger than the non-aggregated other owners.

That's not too surprising, right? Taken together, the vowels cover more of the alphabet than any other single letter. Undoubtedly true, but what does that really tell us?

If the big three tend to vote with management, that sounds potentially concerning. But are they just not voting for wacky shareholder proposals? What proportion of those votes actually tip the outcome?

The section on the potential impact of this ownership concentration has a lot of "could have" and "may be" in it. We should definitely be doing that research, but at this stage the story isn't super compelling.

Cool! Any particular reason why you'd shuffle like that, as opposed to using a circular buffer? With IndexOffset, you've almost got a circular buffer going already.

If you've been an employee at a unicorn and have common shares/options, check out table 8 (page 48) that lists common share valuations according to their model.

AIUI, the common share FMV they list should be comparable to the 409A common share valuations you may have gotten. Roughly speaking, assuming the funding round in the table is close in time to the 409A valuation.

> A city like Fremont was master-planned with enough space for both, but is being forced by the state to build houses instead for people who will work in Sunnyvale which has too many jobs and not enough houses.

The legislation discussed in the article is largely about making Sunnyvale (and the rest of the job-heavy housing-light peninsula cities) build more housing.

The height limit you're thinking of is higher. The IBC allows for five stories of wood construction over a multi-story concrete podium (often one or two stories).

http://www.structuremag.org/?p=10934

The main height limit is local politics around density, not building codes.

Just open-sourcing and upstreaming the KMD would resolve the biggest pain-point in linux. The "very fast implementations of extremely complex libraries, and optimizing compilers for complex programming languages" can be separate.

> Hail a cab? You can even use a cab-hailing app? There is nothing magic about Uber/Lyft's service that a municipal taxi service couldn't use.

Are you arguing for or against ride-hailing apps, here? The original comment you replied to was that ride-hailing (as exemplified by uber/lyft) is a much better experience than existing public transit.

> And an aggregated bus service through Uber isn't going to be any better.

Compared to existing public transit (on a set schedule/route), uber/lyft are better because they're demand-dispatched based on where people want to go and when.

> Young people, old people, poor people, mentally handicapped people, people who have been away from power outlets for a while. Foreigners who don't have domestic SIM cards yet.

Many of those people will also have a hard time figuring out which buses (with which combination of transfers) will get them where they want to go. Assuming that public transit even goes where they want to go in the first place.

You're concerned about excluding people, but you're ignoring all the people who are excluded by transit routes/schedules.

> Bus lines and schedules aren't exogenous. They're put together based on estimations of where the demand for the bus line is.

Those estimates are exactly that, estimates, and the resulting routes are subject to lobbying. Plus, the bus lines don't change very often--if they did, it would be very confusing.

But demand-dispatched transportation like uber/lyft know exactly where people are starting from, exactly where they're going, and exactly when. If there's a popular show at a music club, the bus lines aren't going to adapt to that in real time.

> You call a car and it says you're waiting 15 minutes for the next ride, that's no different than going to the bus stop and seeing that you'll be waiting 15 minutes for the next bus.

The difference is that I'm waiting at home rather than waiting at the bus stop. And that the car is going where I want to go, without transfers.