The Future of Journalism is in Refunds

Dutch jour­nal­ism exper­i­ment, Blendle, on what they’ve learned from their first year of oper­a­tion. It’s a pret­ty inter­est­ing idea: you buy a sub­scrip­tion, read the sto­ries, but if there’s some­thing you don’t like, you can request a refund. What they’ve found is that peo­ple don’t want to pay what they can get for free:

We don’t sell a lot of news in Blendle. Peo­ple do spend mon­ey on back­ground pieces. Great analy­sis. Opin­ion pieces. Long inter­views. Stuff like that. In oth­er words: peo­ple don’t want to spend mon­ey on the ‘what’, they want to spend mon­ey on the ‘why’.

And they don’t want to pay for what they per­ceive as with­out val­ue:

Gos­sip mag­a­zines, for exam­ple, get much high­er refund per­cent­ages than aver­age (some up to 50% of pur­chas­es), as some of them are basi­cal­ly click­bait in print. Peo­ple will only pay for con­tent they find worth their mon­ey. So in Blendle, only qual­i­ty jour­nal­ism starts trend­ing.

They have 250,000 users, and some amaz­ing ana­lyt­ics for how they can grow that num­ber.

The Ordinary Plenty from the Romans to the Web

Jere­my Kei­th has writ­ten anoth­er robust and pas­sion­ate defence of the Web. The whole thing is worth your time, but in it he ref­er­ences some­thing he pre­vi­ous­ly wrote, about the val­ue of archiv­ing pub­lic dis­course, what he calls the ordi­nary plen­ty:

My words might not be as impor­tant as the great works of print that have sur­vived thus far, but because they are dig­i­tal, and because they are online, they can and should be pre­served… along with all the mil­lions of oth­er words by mil­lions of oth­er his­tor­i­cal nobod­ies like me out there on the web.

In that piece he ref­er­ences the mar­gin­a­lia of medieval scribes, but this remind­ed me more of Pom­peii. One of the high­lights of my vis­it to the exca­vat­ed town was to see the per­fect­ly-pre­served graf­fit­ti on the walls of streets and pub­lic build­ings. There are tales of rival­ry:

Suc­ces­sus, a weaver, loves the innkeeper’s slave girl named Iris. She, how­ev­er, does not love him. Still, he begs her to have pity on him. His rival wrote this. Good­bye.”

Sweet mes­sages of love:

Vibius Resti­tu­tus slept here alone and missed his dar­ling Urbana.

And just down­right filth:

Weep, you girls. My penis has giv­en you up. Now it pen­e­trates men’s behinds. Good­bye, won­drous fem­i­nin­i­ty!

It’s not artis­tic or even par­tic­u­lary lit­er­ate, but it pro­vides a much more vivid impres­sion of the peo­ple who lived there than any con­tem­po­rary account can ever match. And that’s why I agree with Jere­my on the val­ue of pre­serv­ing the “unim­por­tant”.

Small Numbers, Huge Changes

In a recent inter­view, Sun­dar Pichai of Google dis­cuss­es improve­ments in the accu­ra­cy of their voice recog­ni­tion:

Just in the last three years, we have tak­en things like error in word recog­ni­tion down from about 23 per­cent to 8 per­cent.

That’s the dif­fer­ence between mis­un­der­stand­ing one word in four, to one word in twelve; the dif­fer­ence between com­plete­ly unus­able, and annoy­ing.

Andew Ng, for­mer­ly of Google and now of Baidu, expands on this:

Most peo­ple don’t under­stand the dif­fer­ence between 95 and 99 per­cent accu­rate. Nine­ty-five per­cent means you get one-in-20 words wrong. That’s just annoy­ing, it’s painful to go back and cor­rect it on your cell phone.

Nine­ty-nine per­cent is game chang­ing. If there’s 99 per­cent, it becomes reli­able. It just works and you use it all the time. So this is not just a four per­cent incre­men­tal improve­ment, this is the dif­fer­ence between peo­ple rarely using it and peo­ple using it all the time.

It’s fas­ci­nat­ing to see how these small num­bers make a huge dif­fer­ence; you might think Google’s 92% accu­rate is only a lit­tle less than Baidu’s 95% accu­rate, but in prac­ti­cal terms there’s a big gulf. And it gives me pause to think about the mon­ey, human resource and com­put­ing pow­er spent on try­ing to make those small huge increas­es in accu­ra­cy.

A conversation with a bot

It’s approach­ing 3 AM on Christ­mas Day in 2013, and a South Kore­an teenage girl who goes by the Twit­ter han­dle @jjong_gee texts her friend, Jun­myun, to con­fess a per­son­al secret. She’s depressed, and she needs sup­port. “There was a man named Osho who once said ‘don’t be too seri­ous, life is like a mov­ing pic­ture,’” replied Jun­myun. “If you treat what comes at you like a game, hap­pi­ness will come. I want to see you hap­py.” The girl tweet­ed a screen­shot of the text, thank­ing him for the kind words. But Jun­myun, with his words of wis­dom, is actu­al­ly not a real per­son. Jun­myun is actu­al­ly a bot pro­grammed inside a pop­u­lar Kore­an tex­ting app called FakeTalk, or Gaj­ja-Talk in Kore­an.

The App That Lets Depressed Teens Text with Celebri­ties and Dead Friends. Every time I read some­thing like this I remem­ber how bril­liant­ly pre­scient Black Mir­ror can be.

Of course, I had to have a go myself:

A conversation with a chat bot which ends with it declaring its love for me

What’s inter­est­ing is how, despite the bot not being Tur­ing-com­plete, I still felt com­pelled to con­tin­ue the con­ver­sa­tion, and became quite ner­vous at the abrupt turn it took at the end. After all, I don’t want to hurt the feel­ings it doesn’t have.