laxatives

This could have easily been 3-4 landmark papers, but instead its packed into one cogent idea.

That common interview question about query autocomplete/sentence completion? Shannon solved it and demonstrates it in this paper, almost a decade before FORTRAN existed. New grads still struggle with that problem. PhD's still struggle with that problem.

Pretty much every machine learning classifier is using a loss function described in that paper.

There's a very strong correlation to Uber/Lyft in the industry. Both are investing heavily in companies like Bird to capture the last mile of transportation. Acquiring companies like Bird that help with that are opportunities to upsell customers for a few dollars more to get into a car.

Its the New Yorker -- every article they write reads they like they get paid by the word.

Google's AI Research challenges are so technically advanced, they can only be adequately described by their leading marketing experts.

Do you keep SIM cards after each trip? Seems like a hassle having 15 phone numbers to track, unless you're fully committed to being on Google Voice or something similar. Why not just pick up a new SIM each time you travel? Or just pay for an international plan if you travel frequently?

You do it. A good chunk of that is copying an alraedy existing product.

Probably a significant chunk of the world's economy is from taking an already existing product and making it more accessible.

String Theory is one of my favorite essays of all time. Discusses the incredible gap between journeymen, low-ranked tennis pro's and the elite players in both ability and quality of life, as well as the gap between journeymen and highly competive amateurs.

Definitely recommend Infinite Jest like probably anyone else in this thread, but would recommend reading his non-fiction first. Jumping into Infinite Jest, I thought Wallace was some wierdo crank for a long time before I "got" his style and what he was going for. Even then, it took a long time before I enjoyed it and he very quickly became one of my favorite authors.

I think Infinite Jest and GEB are two of the only epic books that accomplishes so much, but in two entirely different contexts. They both build intense, well founded and well structured frameworks that lead up to epic resolutions, despite being respectively fiction and non-fiction.

That's pre Bitcoin inflation. Its probably valued over a billion USD now, assuming they kept a decent chunk of the BTC in BTC.

Its a little disappointing to see that half of the comments are people defending their actions to no one in particular, justifying why they act like a jerk but aren't necessarily a jerk.

Its not like being a jerk/sweetheart is some binary or mutually exclusive relation, but it seems like this article struck a chord with a lot of users.

A professor of mathematics noticed that his kitchen sink at his home leaked. He called a plumber. The plumber came the next day and sealed a few screws, and everything was working as before.

The professor was delighted. However, when the plumber gave him the bill a minute later, he was shocked.

"This is one-third of my monthly salary!" he yelled.

Well, all the same he paid it and then the plumber said to him, "I understand your position as a professor. Why don't you come to our company and apply for a plumber position? You will earn three times as much as a professor. But remember, when you apply, tell them that you completed only seven elementary classes. They don't like educated people."

So it happened. The professor got a job as a plumber and his life significantly improved. He just had to seal a screw or two occasionally, and his salary went up significantly.

One day, the board of the plumbing company decided that every plumber had to go to evening classes to complete the eighth grade. So, our professor had to go there too. It just happened that the first class was math. The evening teacher, to check students' knowledge, asked for a formula for the area of a circle. The person asked was the professor. He jumped to the board, and then he realized that he had forgotten the formula. He started to reason it, and he filled the white board with integrals, differentials, and other advanced formulas to conclude the result he forgot. As a result, he got "minus pi times r square."

He didn't like the minus, so he started all over again. He got the minus again. No matter how many times he tried, he always got a minus. He was frustrated. He gave the class a frightened look and saw all the plumbers whisper: "Switch the limits of the integral!!"

There must be some incredible data being mined by Amazon via Whole Foods. At the Whole Foods in Palo Alto, there seems to be an elaborate pricing experiment in the ice cream department. Entire brands (Talenti) are dropped in price 70%. Other times, particular flavors are reduced that much, sometimes less so. For Talenti, the entire brand is sold out. Some of the on-sale flavors go quickly or are also sold out while others are completely ignored, despite the discount (namely anything from Three Twins, who's blurb on their ice cream and their shop at SFO make them sound like assholes). Maybe 25% of the options in that section are discounted, and the amount of discounts varies as well, with a number of different steps like (80%, 50%, 35%, 20%, 10% or similar). Nothing else in the shop is stratified like that. Beer and wine might have 2% of the options on sale. Ditto produce and meat. Most departments offer no discounts.

> The Breitmans had also hired Strange Brew Strategies, a U.S. communications company, to promote their project, and Reuters wrote a news story on May 5 about Draper’s involvement.

> In pitching the story to Reuters, John O‘Brien, a principal of Strange Brew, had made claims about Tezos’ progress. He wrote: “The applications of Tezos, ranging from derivatives settlement to micro-insurance, are real and recognized by industry giants. Ernst & Young, Deloitte, LexiFi, etc. have adopted Tezos in their development environments and labs.”

> On Oct. 3, a spokeswoman for the accounting firm Ernst & Young told Reuters: “The statement is not correct. EY has not adopted Tezos.” A spokesman for Deloitte said Tezos’ code is “one of many technologies we’re considering” with blockchain, but it’s “still early stage and we haven’t used the technology for a client project.”

> Jean-Marc Eber, CEO of the French software company LexiFi, said, “The sentence, as stated, isn’t accurate and unfortunately exaggerated, to say the least.” While there had been “informal contacts,” he said, “at this stage, LexiFi has not adopted Tezos’ technology in its development environment or labs.”

Jesus.

> They claim they had a large team working on the project for 4-5 years full-time.

https://www.tezos.com/static/papers/Tezos_Overview.pdf cites development began in 2014 and with a full staff of 10 core developers. I suppose I misread and was wrongly under the impression that they always had those numbers. But it doesn't change that Arthur and the gang modified their LinkedIn pages after the fact to remove full time employment at their day jobs and replace it with full time employment at Tezos.

> IMO they have no intention of delivering anything...

>> The commits on Github prove that isn't true.

https://www.reddit.com/r/tezos/comments/7791ex/the_path_forw...

Yes development continues, but they've downsized the now unnecessary development team to from 10 to 4 and instead are relying on 3 outsourcing shops. They more or less have to spend the money they raised that belongs to the foundation, which requires significant public auditing and transparency while reducing their own involvement and spending their own windfalls as they please. The entire team except the leadership is gone. And now they are removing the head of the foundation.

> They weren't rude.

You're wrong. Arthur made everything conversation personal and used 4chan style name-calling all the time. Really obscene insults flowed from him almost everytime I heard him speak. He used memes to respond to questions he was probably admittedly tired of answering. But the content was ludicrous. They were Shkreli-like get rich or die trying memes. It might have been tounge-in-cheek, but its impossible to tell with him because he couldn't handle a normal conversation.

They have $200M because people donated. The ICO is described as a fundraiser, not an investment, because it reduces liability and opportunity for government intervention. In exchange for "donation" you get XTZ. Presumably, you expect that XTZ to grow in value to make a net profit.

The upshot is that its a word that is synonymous with investment, but requires less responsibility. On the flipside, they could disappear and your are out of luck because the government won't regulate donations as carefully as securities.

I had been following Tezos closely since November and was really excited about the ICO. But it became incredibly apparent that the Breitmans were dishonest.

Firstly, the compensation structure was ludicrous. Many people were upset about an uncapped ICO -- I thought this was fine, until it became apparent that they would make hundreds of millions of dollars IMMEDIATELY. They have absolutely no incentive to continue the product or work another day in their lives. A huge portion of their new found wealth doesn't have a vesting period or depend on any tangible benchmarks.

Secondly, the background of Tezos is completely revised and modified according to their whims. They claim they had a large team working on the project for 4-5 years full-time. Even a cursory glance exposes that one person worked on Tezos while working full time for a major bank and that the team grew insignificantly 6-12 months before ICO. They even changed their LinkedIn profiles to revert this history.

Lastly, both Arthur and his wife are completely inept socially, or knew they would get rich enough to not give a fuck. I'm not one to judge a person, but its certainly not ideal as a manager and figure-head of a billion dollar project. I didn't even think it mattered until I heard they were incredibly rude to just about everyone in their Slack channel asking the most innocuous questions. I thought it must have been one bad experience. I joined the channel and the very first few words I spoke were met by flaming from Arthur himself. Not a couple hours later, he's raging on another person in the chat room at a time when there might be 100 messages/week. Every interaction he had over the next week oscillated between flaming and get rich memes. Later they would delete channel history and eventually shut them down completely under the guise of "preventing fishing attempts", though I never saw a suspicious link anywhere. Arthur stepped down as President because of these issues, but his wife did not fare much better (although I never interacted with her directly).

The ICO seemed worse than burning money. It would be like giving money to organized crime or the opposition party.

IMO they have no intention of delivering anything, except maybe to save face. If the past is any indication, they don't give a shit about saving face either.

100% the objective is get the most out of experts given the opportunity. Its completely unreasonable to be an expert in everything. But it can be difficult to engage an expert if they aren't there to talk shop. Having the right framework to even phrase a meaningful question often requires quite a bit of context and prerequisite knowledge. If someone tried to strike a conversation with me with vague, malformed ideas and buzzwords, I may respond with a semi-scripted answer because I don't know this person and I'm not going to assume their background is the same as mine.

Its completely unfair, but the way that first question is framed, depending on that field, can demonstrate whether this individual is "in the club" and has sufficient information to carry this conversation beyond the most superficial layer. It would be nice have that fluency in many subjects, but browsing wikipedia is certainly not the way to go about obtaining it.

Yeah I guess that's fair. I guess my point is that when you meet that expert, your contribution to the conversation is typically going to be blurting out <pharoah-name> and <pyramid-name> because thats all you remember from that one read, then more or less passive listening until you change the subject. But probably that's more or less enough depending on your goals. What happens when your expert responds, "No I didn't work on that, I did <x>"? Yeah you can try to engage them, but at that point you may have no more context than before you read the executive-summary.

I would argue that having a more bottom-up learning would help you get more from that next conversation by having a lot more context and opinion then the handful of key-phrases you've managed to recall months/years after. That way, the next time you have access to an expert, you can are building on that knowledge, rather than growing a list relevant of words. To be blunt, its the difference between being a student and a tourist.

Yes, I agree. Everything is instant gratification.

Certainly there are diminishing returns to wanting to mastering a particular subject completely and maybe it is better to be merely "good" at a number of things, but its not really meaningful if the barrier to a particular level of expertise is a couple hours of commitment.

On the flip side, this approach to learning means you know next to nothing beyond the dozen first bits of terminology or buzzwords for each new topic, much less the concepts. The moment you finally meet someone interested in discussing one of these things, its going to be immediately apparent the basis of your knowledge is a brief article/book or a weekend watching videos on youtube. Furthermore, these are all going to be little self-contained treatises without any prerequisites -- they are written for a general audience, so you are completely limited to, as the article describe it, the appearance of mastery to people who don't know better.

What's worse is that you seriously have to question the qualifications of any single source claiming to be a valid resource for a wide range of topics. Even if the individual course is written/taught by an expert, they are paid or motivated by another entity with their own set of bureaucracy and expectations. What works for learning tennis isn't going to produce the same class format as what works for learning Film Scoring (I'm looking at you Masterclass, for producing that joke of a class).

In my mind, the best approach to learning requires you to initially commit to either a top-down or bottom-up strategy. If you want to master the topic, you should go bottom-up; focus initially on theory and gradually more on practice/experience. If you want to ramp up as quickly as possible, go top-down and experience something for a more general audience to get a taste. If you like it, you can keep pursuing that route or commit towards a more bottom-up approach, which provides a higher ceiling albeit on a longer timeline. But don't go to a general resource like the one in the article, go directly to the expert's material, first-hand, for the raw, unadultered version.

Highly recommend Tensorflow for anything embedding related. The Tensorboard UI alone is enough make the switch worthwhile, unless you can develop a similar interface. Tensorboard's graph UI is very handy for debugging/presenting as well.

Don't know why you feel so compelled to give your background while writing an anonymous comment on the internet. It only makes me suspect your argument doesn't hold up without some expectation to respect your authority. All the theory in the world is irrelevant until it you can apply it to predict something new -- its almost trivial to create a theory that describes some phenomenon accurately, after-the-fact. See the tremendous bifurcation in physics with regards to string theory. Similarly there are loads of academics in various fields claiming to have been the origins of artificial intelligence.

Author seems to lack an understanding of what constitutes an equation or a conditional.

For just about everyone, the best thing to do is shut the fuck up and do your job.

Is ProtonMail your only active email account?

Does this mean ProtonMail is more reliable? Would you recommend using ProtonMail?

Probably a large part of that is because they are much more permissive to bots/crawling than Google searches.

Either this is satire, or you've completely missed his point. There is little-to-no hand wavyness in the response; He flat out describes the article as pretentious, bloviating non-sense.

The word aside, he's not wrong. That article is filled with flowery non-sense, making long prosaic paragraphs from thin-air. I like reading long articles, but this doesn't contribute much.

Of all the orgs to have a live demo, I would have guessed Bret Victor's lab to it. Still, looks pretty interesting.