Morfternight #94: Algorithms & Breakthroughs
Morfternight #94: Algorithms & Breakthroughs
The one with a photo of a bird that a computer can recognize.
đ tl;dr
Today, we have a photo of a bird and a video explaining (among other things) how a computer recognizes a bird on an image. Also, Steve Jobs made 40 years ago a point that is crucial today.
𤊠Welcome to the 199 new Morfternighters who joined us this week.
I love having you here and hope youâll enjoy reading Morfternight.
Share with your friends by clicking this button.
đ Good Morfternight!
Greetings from Vienna! Two issues of Morfternight in a row from home; hereâs something that hasnât happened in a while.
I hope you are all doing well, and as a reminder, you can tell me about it by replying to this email. I read all emails and reply.
You can also reply to this email to share feedback with me. Do you like Morfternight? What would you like to be different, if anything?
đˇ Photo(s) of the week
Raven, Vienna â More Photos from Berlin
Vienna has a large bird population, from pigeons to ducks, from seagulls to swans, but ravens and crows are the most beautiful and wisest of all. I hadnât even noticed this one on a car just next to me, but I heard a clicking sound as it moved on its windshield.
đşď¸ A few places to visit
And remember, if you're writing in English, that a lot of your readers won't be native English speakers. Their understanding of ideas may be way ahead of their understanding of English.
This is an excerpt from Write Simply, a short essay by Paul Graham. Great advice on writing to be understood.
Ask vs guess culture is an excellent essay by Jean Hsu about a cultural but also generational gap that we probably all have encountered at home and work.
I truly enjoyed this Geoffrey Hinton interview, although I have to say, the âgodfatherâ moniker is becoming a bit annoying.
The conversation touches on multiple aspects of AI, from a very understandable explanation of backpropagation algorithms at play in neural networks to the potential dangers of AI and some advice to future researchers:
The fact is either you have good intuitions or you don't. If you've got good intuitions, you should listen to them and follow your intuition and work on that until you discover why it's wrong. If you've got bad intuitions, it doesn't really matter what you do, so you might as well follow your intuitions.
âď¸ âIt was a breakthrough because people already knew how to use it.â
ChatGPT recently became the app with the fastest adoption ever, with 100 million users over the first months after its launch.
In this 1984 video, Steve Jobs compares the introduction of the Macintosh graphic user interface and the mouse to that of the telephone a century earlier.
In both cases, a new technology saw mass adoption and disrupted the incumbent one quickly âbecause people already knew how to use it.â
I said it before, this is the real revolution of 2023. Not a sentient machine, but just one that, finally, after all these years, people already know how to use.
The video is long, but I linked to the precise spot where he explains this analogy, and that bit is about two minutes long.
If you donât feel like watching it, and prefer to read, hereâs the transcript.
If you go back about 100 years to the 1880s, there were approximately 20, 25,000 trained telegraph operators in the United States and you really could send a telegram between Boston and San Francisco and it would take about 3 or 4 hours and go through the relay stations and it really worked. And it was a great breakthrough in technology that had been around for about 30, 40 years.
And there were some people that talked about putting a telegraph machine on every desk in America to improve productivity.
And what those people didn't know was about the same time Alexander Graham Bell filed the original patents for the telephone. A breakthrough in technology because putting a telegraph machine on every desk in America to improve productivity wouldn't have worked. People wouldn't have spent the 20 to 40 to 100 hours to learn Morse code. They just wouldn't have done it. But with the telephone, within 10 years there were over 200,000 telephones on desks in America.
It was a breakthrough because people already knew how to use it.
It performed the same basic function, but radical ease of use. And in addition to just letting you type in the words or click in the words, it let you sing. It let you intone your sentences to really get your meanings across.
We are at that juncture in our industry right now. There are people suggesting that we should put a current generation box on everyone's desk to improve productivity. A telegraph, if you will. And we don't believe that. We don't think it will work. People will not read those damn 400-page WordStar manuals. They won't carry around these cards in their pockets with 150 /WZs. They're not going to do it.
And what we think we have here today, finally, is the first telephone.
And in addition to letting you do the old spreadsheets and word processing, it lets you sing. It lets you make pictures. It lets you make diagrams where you can cut them and paste them into your documents. It lets you put that sentence in bold Helvetica or old English, if that's the way you want to express yourself.