1/ Productivity tip: when I keep procrastinating on something I often trick myself by saying, "ok just work on it for the next 20 minutes" or some short period of time, "and then you can chill/relax guilt free" and give myself some treat (sugar, netflix, etc)

An irony of facebook's situation.

Sheryl got her MBA at Harvard. One of the most famous cases (Extra Strength Tylenol) in one of the most famous classes (Business History) she took: in 1982, someone put cyanide in Extra Strength Tylenol capsules and killed 7 people in Chicago.

What do you do when someone turns your product into a weapon? When they use the system you built to harm? James Burke, CEO of J&J, was shockingly open with the public, he pulled the product and made significant packaging changes to make product safer (but not tamper-proof).

He over-shared every step along the way re investigation, redesign, stood up as both CEO and human. The reintroduction of new Extra Strength Tylenol succeeded. Burke saved the brand.

But four years later it happened again. A killer put cyanide in the capsules, this time a woman in Yonkers died. Same CEO, Burke, pulled the product again, completely changed the form factor from capsule to caplet and relaunched *again*. It worked *again*. How'd they do that?

Burke (CEO) tapped J&Js goodwill bank account w/ the public. Two big withdrawals from that bank account in four years + 8 dead bodies! But his honesty, openness, humanity (choked up about the deaths more than once), humility kept the goodwill bank balance positive the whole time.

Can we control the output of neural generation systems?
Can we build a better chatbot using this control?
And what makes a good conversation anyway?

Check out the blog post for our NAACL paper:

https://t.co/UAAUxSeA0L

#AI #NLProc #DeepLearning #ConversationalAI @facebookai

This post got a little delayed since the paper's publication, but the intervening time has given me an opportunity to reflect.

So this post has some extra thoughts regarding more recent NLG developments - e.g., pretrained models like GPT(2), and the suitability of beam search.

Originally, tweet length was 160 minus username length. One day, I turned to Jack and pointed out that I got one more character than him. He said, "That's not fair. We should standardize." I agreed. He said, "How about 140?" I said, "Sure." And he updated the code.

Last year, we upgraded tweet length to 280 and guess what? Use of 'please' is up 54% and 'thank you' is up 22%. I love that! Also, people are asking more questions and having more conversations. All this, and the majority of tweets are still under 140. It worked.

Many readers have asked me "why do so many pulp covers feature women in ripped red blouses standing in swamps while a man fights off an unusual animal attack?"

The answer is artist Will Hulsey...

Will Hulsey was the undisputed king of the animal attack pulp cover. You name it, he'd paint it attacking you in a pool of stagnant water.

Very little is known about Will Hulsey, but he worked on a number of men's pulp magazines in the 1950s and early 1960s including Man's Life, True Men, Guilty, Trapped and Peril.

Their audience was ex-GIs: during WWII the US Council of Books in Wartime had given away over 122 million books to American servicemen to read; this led to a post-war surge in paperback and magazine sales amongst these newly enthusiastic readers.

As a result the 1950s saw a raft of men's pulp magazines being published to tap into this market - almost 200 different titles!

It’s underremarked upon that stablecoins accomplish basically all of the original design goals for Bitcoin qua a transactional mechanism (instant free value transfer anywhere), but that nobody cares about this because nobody ever cared about that part of the pitch.

“Why do people care about stablecoins then?”

A mix of “they encourage dollar-denominated liquidity in the cryptocurrency ecosystem and discourage withdrawal of the same” and “they’re good for money laundering.”

“But they make value transfer between exchanges much faster!”

This was a solved problem in traditional finance, too, mostly through the extension of credit. (It doesn’t matter how long settlement takes if there is sufficient trust to enable credit.)

The Bitcoin ecosystem is *positively allergic* to credit, so you have to call it a coin for them to accept it. And after you call it a coin they ignore everything the world has learned about credit, like risk management.

“Stablecoins aren’t credit!”

They’re pretty much exactly credit? A tether is a zero-coupon Bitfinex bond with a non-functioning call option. I