BlackJack

ANA is good but isn't even the best airline in its' own country. LH isn't the best airline in its' own group (Swiss and Austrian have better seats/service, and Swiss has equivalently good coverage).

United is probably the best US airline for international but domestically Delta is arguably better.

The British were engaged in the slave trade, then worked to outlaw slavery but replaced it with indentured servitude that was basically like slavery with a trivial income. That and exploitative colonial government meant you don't need slavery to loot everything.

Reparations are a different topic and wouldn't necessarily solve the problems of slavery/colonization.

Processed foods usually refers to packaged snacks, food with lots of added ingredients, corn oils, artificial ingredients, etc or fast food/junk food ordered at restaurants/fast-casual places.

In the past, he had call options that he could've sold for profit but instead he exercised them to get even more GME stock. It's literally the opposite of dumping.

The whales manipulate markets much more than anything Keith Gill has done.

Disclaimer: I work at Google on related stuff

Veo is the name of a video model. VideoFX is the name of a new experimental tool at labs.google.com, which uses Veo and lets you make videos.

Thanks for the feedback though, I see how it's confusing for users.

Guaranteed value is a licensing payment that compensates the publisher for allowing OpenAI to access its backlog of data, while variable value is contingent on display success, a metric based on the number of users engaging with linked or displayed content.

...

“The PPP program is more about scraping than training,” said one executive. “OpenAI has presumably already ingested and trained on these publishers’ archival data, but it needs access to contemporary content to answer contemporary queries.”

This also makes sense if they're trying to get into the search space.

That's not even close to true. Look at the massive growth of slavery and religious violence under Arab and Muslim conquest across Asia and Africa.

Arab and Western conolialism have different flavors but enormous exploitation and abuse.

Archibald sells a vicuna sweater for $3.1k, and they say the at cost price is $1.9k [1]

Given maybe $500-800 in production cost, shipping, duties etc, the cost is probably $1k. It is significantly more expensive than other fibers.

1. https://www.archibaldlondon.com/us/collection?category=vicun...

So sad to hear that. QDB brought me so much joy to read.

Love this book, and looks like stripe.press has some other really good books like Hamming's The Art and Science of Doing Engineering.

That being said, the UX of this page is horrible. Scrolling hard, navigating hard, bad font. I cannot believe they put this out.

I've read most of Matt's "Notes on [X]" In general they are well written and interesting. In some cases though it's too simplified and at times insensitive. In particular, I've talked to people who are from Nigeria/Ghana or have visited there and they pointed out issues with his narrative. But most people who read Matt's blog have never been to these places, so they take everything as 100% accurate.

That being said, if you treat it as literally just his experiences, then they're fantastic reads.

P.S. His notes on Mauritania [https://mattlakeman.org/2023/08/16/notes-on-mauritania/] is really interesting for the train story. I had never heard of such a thing before.

Have you done the math, even at a ballpark level? Commercial Real Estate (CRE) is estimated to have ~$500B loss in 2023 and nearly the same in 2024 (different from debt repayment) [1]. Stock indices were fine. In general, there's way more private investment/ownership in CRE. The market can suffer a lot more in losses before it starts to meaningfully affect mom&pop. Even stocks like SLG and BX have done fine last year.

[1] https://finance.yahoo.com/news/commercial-real-estate-values...

That's a good point. I agree it's not clear cut one way or another and we gotta let it play out.

disclaimer: I work on GenAI at google, but views are my own

The question is, how did the model create Mario&Luigi or Scrooge McDuck without training on copyrighted data? It can't just crawl Wikipedia because Fair Use in Wikipedia doesn't constitute Fair use for a commercial AI model.

One possible outcome is more transparency on what datasets were used to train the models.

After reading your comment, I checked their history. The account has 7k+ karma, and if they're professional marketer they are promoting:

* DALLE * Stable Diffusion * Tabletop RPGs * Sega CD * Nice suburban housing * Neuralink * Cliffnotes

etc etc. It just sounds like a normal person talking about things they like or interests.

A good PM doesn't have to know the differences, but should understand that mixing encryption protocols is going to break things. Ideally an eng lead would be flagging these issues.

"On June 9, the day of the alleged poisoning, the Florida Department of Business and Professional Regulation conducted an inspection at the restaurant and conducted a follow-up inspection on June 13.

The department found a total of 31 violations at the restaurant, including employees switching from preparing raw to ready-to-eat food without washing their hands and the restaurant was reporting with an expired Division of Hotels and Restaurants license, the PNJ reported."

Hard to feel bad for the restaurant if employees are making raw food without handwashing...

I agree but want to correct you on the public school point. Excluding the service academics, there are 2 or 3 schools with sub 20% acceptance rates, and every other public school is above that.

See [1] and [2]. And these are the best public schools. The average of the top50 public schools is probably 40%ish, and the average of all public schools is 70%.

[1] https://www.collegesimply.com/colleges/rank/public-colleges/... [2] https://www.oedb.org/rankings/acceptance-rate/

I read the comments and many are in this vein: "Agreed, moderators with a power complex are the worst thing about this site. Don't even care about this API crap." Never realized mods were so disliked.

I thought the blackout sent a message+got coverage, but don't like subreddits extending it without consulting their users. I use reddit about 5-10 minutes/day so I don't care either way.

Madoff was arrested fast because his sons went to prosecutors and admitted to the fraud. That's very strong evidence + direct witnesses vs. here where they have to prove fraud and gather evidence.

Amazon Ads 31B is an annual figure. Last quarter it was ~8B. Same thing with cloud; AWS pulled ~18B last quarter.

Amazon Ads is a legit threat to Google, especially with product searches. Instagram struggles to monetize commercial queries to the same degree as Goog/Amzn.

I upvoted this comment, but people are not so dumb to only eat junk food. There are environmental/psychological factors why people overeat. Other research shows that people will overeat "healthy food" i.e. huge dollops of olive oil, or fruit juice, because it's "healthy."

I agree that reducing bad food and portion control is the answer, but it's very hard to drive behavioral change in people. And the health industry has done a "good" job of creating medication to enable bad lifestyles i.e. metformin for Type2 Diabetes, other drugs to manage cholesterol, etc.

I like the strategy and employ it on a portion of my 401k. I think you need to do it in a tax-advantaged account, otherwise rebalancing + short term gains will eat away your profits.

Also, that strategy fails in a rising interest rates environment like in the 50s and 60s (not sure of exact years). Fed has indicated keeping rates low for the next two years, but if they start hiking rates after, I think the strategy would underperform.

Thanks for the population numbers, I didn't know they were so different. I just ranked them from (my perceived) livability - SLC close to skiing/nature, while Minneapolis/Ann Arbor are not destination cities like the other ones.

I think Seattle / Boston can be T1, but then you'd have too many T1s. My logic was "do lots of people want to move to this city?" and I think the answer is "no" compared to the demand for the T1s.

I see your logic though, and maybe it's better to just have a T1 and T2.

Props for posting an ordering :O

This seems to be ranked on eng availability, but in terms of actual city tiers it's more like

T1: SF / NY / LA / Miami / Chicago / Atlanta T2: Seattle / Boston / Austin / SLC / Portland T3: Minneapolis / Ann Arbor

LA is one of the densest parts of CA though, outside of SF/Bay Area. So if LA is 3%, you could assume the state as a whole is maybe 1-2%. Then 17x that becomes 17-34%, which seems more reasonable.

Right from the CEO's mouth :-)

This will be great for students, thanks Matt.

Please email me (email in profile) and I will try to get you support internally.

Disclosure: Am a googler, know nothing about any of this other than public press

I laughed out loud watching it live and reading the quote later:

"I want to thank Google. Google is helping to develop a website. It’s going to be very quickly done — unlike websites of the past...Google has 1,700 engineers working on this right now. They have made tremendous progress."

"1700 engineers!" :-)

Congrats on the launch! This is definitely a real, big problem.