arciini

Summary (self-written)

Factor 1: vertical integration and different supply trains

- US automakers are really just car assemblers that build the engine and frame, and assemble everything else

- EVs are uniquely good for vertical integration, and Tesla recognized this when the Roadster had lots of issues. They gradually integrated, but still relied on external partners for earlier models' interiors and key parts

- Now, China's EV makers have taken this to the end

Factor 2: US trusts the private sector too much, and they make short-sighted decisions

- In the US, mines and refineries for raw materials have largely closed due to low margins

- Apple considered going into EVs, but they decided not to due to lower margins and now have a movie studio

- In contract, China ordered state-owned enterprises to do the mining/refining and build the charging infrastructure that's not lucrative

Decision that needs to be made: what level of protectionism the US should have:

- Japan and Germany both protected their local carmakers so they can catch up and (ultimately) exceed the success of foreign ones

- The US carmakers only exist due to the tariff on foreign trucks

- Welcoming German/Japanese carmakers to create plants in the US allowed us to get good products

- China's EVs kinda sucked despite government subsidies, so they invited Tesla around 2020, which set the gold standard and also forced their local suppliers and workforce to match Tesla's standards. Those suppliers then supplied the other new EV carmarkers

- I think what the video is trying to say: we should also encourage the Chinese carmakers to create plants in the US to give our EV industry a shot in the arm

I'd guess that this is to ensure one abusive user doesn't get every other user blocked from a large service (say, Google) for botting over the VPN and constantly rotating IPs.

It's a practical measure, but definitely has a privacy cost though.

I was pretty convinced by this article to not use TIOBE as a mark of a language's popularity: https://nindalf.com/posts/stop-citing-tiobe/

Its primary point is that TIOBE is based on *number* of search results on a weighted list of search engines, not actual usage in Github, search volume, job listings, or any of the other number of signals you'd expect a popularity index to use.

It could easily be indicating that Python articles are being generated by LLMs more than any other class of articles.

Hi! Co-founder of Wanderlog (YC W19) here - what issues did you run into with it? I'd love to get your take to see if there's anything we can do there.

For me, definitely! It helped us raise our first round of funding, without which we may have failed during the pandemic. I also learned a lot and found the online platform useful.

I did YC when I was 27 after trying to do a tiny bit of fundraising, and kinda agree with you, but I think you can replace "22 year old" in your comment with "anyone not super-well-networked in the startup scene".

The truth is, even if you have pedigree (top-tier college, worked at FAANG and startups), the very first time fundraising is going to be pretty hard and annoying unless you've already been going to a lot of founder meetups, want to start something in a hot field, or have VC's in your personal network. This is especially true after the money wave from pre-2022 receded.

I do feel like I probably could've raised on better terms, but I'm not sure it would've improved the trajectory of my life or expected value. YC taught me a lot about fundraising that I didn't know I didn't know. Compared to other VC's we've worked with (who have all been pretty great), YC has provided more specific advice to our stage, better technology, and a slightly better network.

I personally also strictly disagree with their feedback: "try many small things, experiment, tinker, and build a portfolio of multiple income streams".

This is very much the indie-hacker route, but honestly, if you have one thing that's working out, it's usually better financially (though maybe less fun if you like purely doing engineering/product work) to try to grow it than to build a portfolio.

As someone who went through YC, but started out more on the indie-hacker side, I want to mention that going through YC in particular does not close the door to just continuing to build a successful, growing business without raising one round after the other.

I think some indie-hacker influencers encourage this kind of us vs. them thinking about VC's, and the truth is somewhere in between.

I know of many businesses who have decided (either by choice or through a lack of fundraising options) that the ideal way to grow and scale their business after YC is to not fundraise. It's something that YC partners explicitly acknowledge. They understand that an alive business can continue to grow, whereas a dead business is just dead.

Given there are 3 credit bureaus, is there a way to avoid having a credit score at one of the credit bureaus? I think that's a way that we as consumers could try to increase competition in the field.

I did some Googling and it didn't seem like there's an easy option.

While I agree Google is being a bit sly in lobbying against privacy legislation, I think they have a legitimate point, but also think that it'll lead to them concentrating more power in the ad market.

Google clearly believes that they can obfuscate identity by enough so that ads can be targeted while privacy can also be preserved.

If you don't think that all ads are bad (and Google argues that ads help make many of the sites we rely on economically feasible), then I think you may agree that sometimes, ad targeting produces more useful ads. Many of my friends have actually found Instagram ads interesting enough to talk about them at dinners. It's better to have those ads rather than random banners for things that you don't care about. Google would definitely also argue that ad targeting improves the worth of the Internet and allows more sites to offer their services for free because they can make more from AdSense.

The following only makes sense if you buy this argument:

The data for ad targeting has been abused so often that for many (most?) consumers, it's not worth it.

Google's perspective is: "we can be a responsible steward of this data for this new age of privacy-conscious ad targeting". The Chrome topics API and mathematical/statistical obfuscation are things that a blunt tool like the law may forbid. As far as arguments go, I think this is actually a plausible one. I do think Google has somewhat OK privacy controls compared to other large tech companies, and way better ones compared to bad acting small sites and ad companies/data brokers.

That being said, I don't love the concentration of power (that's why the DOJ is going after them) - I'd much rather there be some decentralized way to ensure privacy but still allow useful ads, but we get what we get.

While I'd love to attribute this to data harvesting and don't love Meta as a company, as someone who has tried to scrape Instagram in the past (to get recent images users have posted for specific restaurants), I believe this is a reasonable measure to increase the cost of new spam accounts.

The cost of an email is virtually 0. The cost of a unique phone number that can receive text-messages is non-zero. This was a pain in the butt for me, as you often have to get a real phone number (they reject VoIP ones) and that takes more work to get working.

Bots and scraping is a huge issue for Instagram. Instagram really dislikes the fact that you can buy a lot of likes for very cheap, so I kinda understand why they do this.

It seems like this is mostly summarizing what seems to be pretty settled law. Here are my main takeaways:

1. The courts have concluded that just because a price is higher does not mean it is necessarily in the fiduciary interests of shareholders (especially long-term ones). This is reasonable too, as in many ways, shareholders are more shortsighted

2. Even in the short term, shareholders are usually incentivized to accept offers, whereas auctions and negotiation may yield better offers (this reminds me of the latest season of Succession)

3. Somewhat interestingly, a study (Kidder) showed that companies that defended takeovers successfully were worth less, but companies that tried to defend a takeover but ended up being acquired by the original buyer or another buyer ended up being worth more

4. Different structures of offers (a 2-stage offer, where 51% of shares are acquired at one price and the remaining is acquired at another or not acquired) can make the offer pretty bad to a lot of shareholders, even though it coerces shareholders to pass it

The article concludes that because of all these factors, the board having more power is generally a reasonable thing. I think this generally matches up with the overall role of boards, and I generally agree.

Our analyses by testbed instruments that are on or will be sent to Mars unveil that although the mineralogy of Red Stone matches that detected by ground-based instruments on the red planet, similarly low levels of organics will be hard, if not impossible to detect in Martian rocks depending on the instrument and technique used. Our results stress the importance in returning samples to Earth for conclusively addressing whether life ever existed on Mars.

This study seems to support the Mars Sample Return Mission, which has been controversial lately due to its high costs: https://arstechnica.com/space/2023/07/the-senate-just-lobbed...

The concern expressed by some scientists, including former NASA science chief Thomas Zurbuchen, is that the ballooning cost of Mars Sample Return will cannibalize funding from other science missions. And if the price is already approaching $10 billion now, then it is likely to spiral further out of control.

While I don't love this API's idea, I understand why they're doing it, and the API it describes really just sounds like any Captcha API today.

Google's plan is that, during a webpage transaction, the web server could require you to pass an "environment attestation" test before you get any data. At this point your browser would contact a "third-party" attestation server, and you would need to pass some kind of test. If you passed, you would get a signed "IntegrityToken" that verifies your environment is unmodified and points to the content you wanted unlocked. You bring this back to the web server, and if the server trusts the attestation company, you get the content unlocked and finally get a response with the data you wanted.

The problem with Captchas today is that there are a lot of services you can use to bypass them. You send the token to a human, human gives you the solution-token, and you pass that to Google.

I can see why they want to make this more protected. As a user, if this lets me solve captchas less for certain sites, I'm OK with that. Of course, I don't think this API should be used for the entire web, but I definitely understand its use-case.

I think this is actually a pretty balanced and rational take on crypto. There's a lot to be jaded by in the industry, and I'm glad he got a lot out of the experience (attending the conference, meeting people, including other skeptics) despite not continuing on the startup. I think at the start of our careers, it's generally a good instinct to just try things and see if they're for you.

By the way, we didn't plan doing that at the end, after thinking about it a lot, but it meant we would need to gain many users by having a killer product and work for free for very long, like a regular startup I'd say.

Real web3 startups are just like normal startups if you take out the token economics. There do exist use-cases and users, but I'm glad the author had a strong enough sense to do what he felt was right. It's definitely easier now that the bandwagon has mostly stopped, but I think his reaction is a good one to maintain throughout our careers.

The article mentions that:

Stratospheric SO2 rains out of the atmosphere after a year or so, so unlike CO2 emissions it is relatively easy to control the dosing and the magnitude of the desired effect

Is this not true? I'm curious about where I can read more about research about the lifetime of atmospheric SO2.

Edit: The best source I could find about stratospheric SO2 was based on measurements after Pinatubo at

https://www.osti.gov/pages/servlets/purl/1439705

(https://doi.org/10.1002/2017JD027006)

I see that the concentration declined to roughly 1% after 180 days (Figure 7). Volcanic aerosols seem to last a bit longer but seem to go back to background around 4 years (Figure 10).

Housing as a for-profit business kinda makes sense, but I think the main takeaway of the article is this:

The thing is, "the algorithm" should have very little to do with that sick feeling. The coldness of the interface and robotic voice certainly make for a stark contrast with the thing you are doing, but they aren't the cause. The moment you start thinking about someone's longtime home as something that can "align with strategy," and about pricing someone out of their longtime home as "an adjustment that would be more beneficial," you have morally lost. It doesn't really matter how you go about making that adjustment.

The tool and the training for the tool definitely matters, but the truth is, making business decisions about housing will necessarily hurt people in ways that are disproportionate to the amount of additional income a landlord gets.

This sucks for everyone, but especially for less mobile or financially-secure tenants. I don't think there's a great solution for the American rental system yet other than less profit-maximizing owners or schemes like HDB flats in Singapore.

This thread leaves a lot of unanswered questions:

1. This was likely mitigated through a device update. What version did it roll out with? Which devices are still unpatched?

2. How was it compromised? Was it an OEM? An internal leak at Google?

3. What is the attack vector? It sounds like it was likely side-loading apps used by some attacker, but did any of these make it onto the Play Store?

I'm sad that this is happening to an app that's useful to its users, but the reality is that scraping is legal, always possible, but difficult.

This particular case is a bit harder since it's not purely using public data, but may still qualify since it's likely scraping with legally-obtained credentials.

I know of businesses (scraping for ride-sharing, scraping for business intelligence for retailers, scraping from LinkedIn - see HiQ Labs v. LinkedIn) that have continuously succeeded via scraping in ways that large businesses oppose.

The key is: you must make enough profit to justify dedicating engineering and legal techniques to defend your scraping.

- Scraping public data is legal, as affirmed by the Supreme Court in Van Buren v. United States [1] and HiQ Labs v. LinkedIn [2]. Defending yourself or suing the data owner in court are both expensive though

- Defeating anti-scraping via technical means is pretty much always possible, but can be costly depending on the scraped site's technical expertise and value in keeping their data private. The benefit to you must exceed the cost to you, and ideally should also exceed the cost to the data owner

- Mobilizing PR and internal resistance may also be effective, but it's usually hard to have outcry from a large enough group to change an organization's policies. In this case, the union can push for it, but AA may try to withhold improvements until the next set of union negotiations

1. https://en.wikipedia.org/wiki/Van_Buren_v._United_States

2. https://en.wikipedia.org/wiki/HiQ_Labs_v._LinkedIn

This seems like a perfectly reasonable way to end a service. The server died, and the effort to bring it back up seems high.

This is a reminder to pay for services you depend on. Per the Pinboard founder's post:

I love free software and could not have built my site without it. But free web services are not like free software. If your free software project suddenly gets popular, you gain resources: testers, developers and people willing to pitch in. If your free website takes off, you lose resources. Your time is spent firefighting and your money all goes to the nice people at Linode.

Like a service? Make them charge you or show you ads. If they won't do it, clone them and do it yourself. Soon you'll be the only game in town!

https://blog.pinboard.in/2011/12/don_t_be_a_free_user/

It's fascinating that the article ends with a very pessimistic quote, but the implication is clear: wiretapping was immune to policy solutions, but technical solutions have been far more effective at solving this problem.

Futility was the order of the day. “Most experts believe that no matter what legislation is enacted, the unhappy outlook as of now is that wiretapping is here to stay and will increase,” Newsweek reported in an article on “The Busy Wiretappers” in the spring of 1955. The tumultuous decade that followed proved all of the predictions right.

Public-key encryption has brought wiretap-resistant communications to the mainstream through the Internet in a way that would've been basically impossible to do at scale in the analog world.

This is really pretty impressive. I think Copilot for these kinds of one-off analysis tasks where specific data manipulation rather than structuring abstractions makes a lot more sense. Structuring libraries or building UI requires a lot more understanding of potential users - in that case, writing the requirements is honestly the harder part.

As someone who has never really understood monads (or tried that hard to) but who has done a good amount of Javascript programming, I really like the description of a monad as a series of nested data structures describing the next steps of a project.

Ultimately, that's very similar to what old-school ways of declaring Promises do in Javascript. You're creating a data structure that you then attach a new function to execute with the results.

This looks great, but I think it'd be great if you submitted this (if you haven't already) as an issue to Jetbrains! They may integrate a similar fix and that'd be much more discoverable

In Pan & Schleifer (1996), subjects experienced more arm discomfort/pain/fatigue the more keys they pressed while doing a data entry task.[23] Finally, Feng et al. is a recent (2021) cross-sectional study that found that “prolonged computer use time and working without breaks were associated with presence of wrist/hand symptoms”.[24] But of course correlation is not causation, and there are many potential confounders here, e.g. maybe people who spend lots of time at the computer exercise little, and it is the lack of exercise, not computer use, that causes problems.

I generally agree with this. I've found that I have less wrist pain now that (1) I've gotten a better mouse and (2) I work out my wrists and forearm occasionally. I think the keyboard is a much smaller cause, especially since a lot of repetitive stress comes more from gaming and use of mice for me.

Be careful where you store your money. One of his puppies (obviously Sam kept several for himself) ended up digging up one of his stockpiles and started eating his money.

I imagine this isn't a common problem, even for North Korean startups, but was funny to read nevertheless.

Interestingly, there's a whole generation of startups that help startups store their free cash better than normal bank accounts in the US nowadays. The markets!

I think it's reasonable for regulators in the long run to rein in these cash cow businesses a bit, when they're a crucial bottleneck in an ecosystem, just like long-duration patents or copyright. The extra fee likely stifles innovation, just like any tax or overhead to setting up a business.

Both Android and iOS are not actually that dependent on App Store revenue to stay afloat, and I think courts will (and should) continue ruling this way in more consumer-friendly locales.

Apple still makes more of its iOS profit from device sales, despite its attempts to increase services income. Google actually makes more of its profit from ad sales to users on Android phones than from Google Play Store itself. The development cost of Android is likely quite a bit lower per user than what Google has to pay to Apple to keep Google as the default search engine. As such, Google still has a strong incentive to keep developing Android and promoting it so it gains market share from iOS.

Similarly, I think powerful people have always behaved as poorly as in the examples he mentioned. It's just that nowadays, data-driven traditional and social media uses the outrage to make money. This does unfortunately lead more people to think that cheating is OK, but I think for the most part, our systems do eventually work to correct that.

Sorry about that - I don't think that comparison made sense and it was also really inflammatory so I removed it

That’s because I forgot one essential reason bitcoin has survived:

Because people really, really, really want it to.

My impression is that Bitcoin (and crypto) is now a religion. It's tied to a lot of people's identities. That's to say, a group of people think of themselves as crypto people and are proud of it, just as someone might be a Christian and proud of it or a Golden State Warriors fan and proud of it.

While I've personally used crypto and found it useful, it doesn't really matter what the environmental impacts are or how useful it is to a large swath of those interested in crypto. As long as there's a plausible argument for why crypto is good, it's enough for the believers to continue believing.

Not only that, crypto is and will continue to be cool and rich. As long as people like Tom Brady and Fortune 500 companies are talking about buying and selling crypto, it's a religion that will continue growing.

By acknowledging this, I realized that I should move past dismissing the religion, but rather think about how I should engage with it. Imagine I've never watched American football but a lot of people I know are really into it. I then need to think:

Is it a cool community that I should join? Is it something that's not for me but that I should try to understand? Or is it something that's so bad that I should try to stigmatize it? I think the right response for me is closest to the second option, but I'm not sure yet.

[edited to remove some earlier, more divisive comments that I didn't intend to make]

I loved Marcus's example about how he was successful with Puzzle Storm:

I had been studying chess for a long time, and discovered a new mode on Lichess, puzzle storm. It’s a timed mode where you solve as many puzzles as you can in about 3 minutes. It was fun and lined up with a skill I was learning, so it ended up winning in whatever calculus my subconscious does to decide how I’ll spend my free time. This was a nice fluke, but there’s no chance that coding on a side project, for example, will ever win out in in an immediate gratification battle with the time-wasters. So I don’t see how it could be the path to a general solution.

I think in general, the key to "wasting" time effectively is to find high-reward time wasters and make them as convenient as possible.

For me, this has meant selectively blocking websites so that you aren't fully out of options, but so that you read long-form magazine articles instead of Reddit. For others, it could mean having an interesting series of Youtube videos on how to engineer spaceships or on how to apply a new technology you've been curious about.

In times when we have foresight or hindsight, we can bank that and take concrete steps to nudge ourselves towards better choices.