Last updated by at .

AI optimerade vinstmaskiner

Var har jag sett detta tidigare? I alla fall, vilka delar i detta är bekant?

Två bra frågor när vi hör något nytt, ställs inför något okänt och måste hantera förändring. Sällan är allting nytt och 100% jobbigt. Vi känner ofta igen delar. Inget nytt under solen liksom. Fler närbesläktade frågor förfinar det hela, men det är ett annat ämne.


Läste NY Times artikel, i en serie av många, om sökandet efter grundläggande regler för säkerställandet av viss etik inom AI nyligen. Gruppen kom fram till en ganska stor uppsättning. Exempelvis, och föga förvånande, transparens kring företags användande av AI, säkerställandet av diverse utvecklar-team, industrisamarbete kring regler och regelefterlevnad m.fl.


Förhoppningsvis kan en ganska liberal hållning till AI och etik funka – vinstdrivande företag ska sätta egna “checks and balances” på plats. Men kan vi leva med “förhoppningsvis”? I en annan artikel utrycker Brad Smith, Microsoft’s president and chief legal officer “We don’t want to see a commercial race to the bottom…law is needed”.

Var känner vi igen detta ifrån?


Lyssnade nyligen på en AI-podcast (det blir ett par) där Stuart Russell i en diskussion om AI och framtiden krasst konstaterar att (fritt översatt från engelska):

AI-system redan tagit över i världen. De kallas företag (corporations), vinstmaskiner som optimerats just mot vinst.

Good point. De är väldigt lika logiken i AI-algoritmer och utrycker lite av de faror de smarta människorna fnular på för närvarande. Vad händer om de optimeras på ett sätt så att de frikopplas från “vett” och “kontextuellt hänsynstagande” (läs: miljö, världen, välmående på lång sikt).


Intressant liknelse. Inte perfekt. Intressant.

The threat of artificial intelligence more intelligent than ourselves

Detta med artificial intelligence va? EU-parlamentet släppte sin robotrapport. Massa möjligheter men också hot. Kill switch? Jo, det känns kanske tryggast så. DI Digital lyfter även denna biten:

Men utredningen belyser också att robotar kan utgöra ett potentiellt hot mot människor. Det finns en möjlighet att AI kommer att vara vida överlägsen den mänskliga intelligensen, heter det. Om det inte kontrolleras riskeras människans överlevnad.

från DI Digital

Jag slås av att vi också kan våga se det som så att om robotar blir mer intelligenta än oss människor så kanske vi har en bättre chans att överleva. För så intelligenta beslut har vi väl ofta inte tagit? Just a thought.

future of key societal institutions

So, as mentioned earlier, banking is one of the most exciting industries from a service design perspective and definitely one on the verge of really being disrupted. Perhaps less likely by new entrants surprising the old dragons, but rather one or more old dragons to finally redefine and rewrite their ”normal” approach to banking.

In my feed, these two initiatives surfaced. Don’t know how serious they are with this, but we’ll see.

nordea innovation challenge

Nordea Bank Challenge
A team from IBM, Nordea and Apple were available for all the teams through out the event.

Deutsche Bank, future of banking

Design boom - future of banking
A future of banking event by designboom, in collaboration
with Deutsche Bank.

If banking, as a key institution in society yet until recently quite unlikely to progressively move forward, is doing it , who aren’t, yet? Library and other public services? Governments as such? Traditional education system? What area or industry is the least progressive and change embracing, I wonder.

Robinhood and the disruption of financial services

The financial service sector is one of the most interesting ones in terms of burgeoning disruption. A sector up until fairly recently not seen as especially dynamic – mainly because the sector itself seemed, and perhaps still is, largely uninterested in advancing things – is now rattling, shaking, squeaking and bustling in every way.

Simple Bank sold to BBVA. Tink, a Swedish company, raises more capital and claims 2% of the swedish population holds an account. In the same area, that of personal finances and money management, is another Swedish start-up in Dreams, helping individuals better save money.

This kind of stuff is natural when you think of the meaning of banks (or whatever we’ll call it). Why individuals need them in their lives. Entrepreneurs focusing on user value, solving problems and seing unmet opportunities, create these things because they should exist. They make sense. And as living our lives involves money and financial services, and technology allows for it, that industry is now booming. Great for almost everyone.

From private finance to investment and trading

Many of the new players in financial services have focused on payment and personal finance. At least the user facing ones. Robinhood takes it in another direction, that of trading.

If I were part of the old school trading industry, I’d be paying close attention as the waves of disruption come rolling in. But, what struck me in the Wired article on Robinhood, is the quote below, where the founders paint a picture, a usage scenario, that really helps us see the viability in the service, and hence potential democratisation of trading.

We all know how toilet breaks, queuing in line, bus rides, ad breaks, just-after-eating-up, micro breaks etc and so on are devoted (more by some) to not only instagram and Facebook – the status check of the collective also known as group of friends – but candy crush, poker, casino and other game like apps, providing that micro fix of excitement. Every gambling company I’ve come in contact with knows (and struggles to perfectly articulate) that intersection of fun/excitement and risk/benefit tickle. But it’s right there in the middle, which is why trading – when made accessible like this – might very well find its way into that usage occasion. I think Roobinhood perfectly exemplifies the importance of looking at behaviours in combination with technology and business vision.

Standing in line for coffee may seem like an awkward time to trade stocks. But for the makers of the new app Robinhood, those casual moments are exactly when they want to reach a new generation of potential investors who might otherwise feel the markets are closed to them.

Robinhood - democratizing stock trading
Screen grab from Roobinhood

From 2009 to 2014 – the deal-breaker that is APIs

an rfid and twitter powered cat door via psfk.com

Back in 2009, the above twitter/rfid cat door surfaced in my feed. I remembered that me and a colleague of mine (a brilliant digital director now at North Kingdom) had an interesting discussion about possible futures of APIs in general and connected physical things more specifically. It’s really ever since (and before that) that I’ve been bookmarking and in other ways saving and reflecting thoughts on what I label “offline/online”.

If the internet as such is now finding its way into the minds of the less technically advanced, the second revolution re: internet is the physical one.

I often get “oh shit” reactions when I let my thoughts wander. I had one while preparing a deck and a few workshops around the subject of brand strategy and digital strategy considering the new reality that is networks and the dramatic impact it has on businesses, communications, great ideas etc. I had completely missed the Watson API. I mean, Watson computer. Cognitive powers. AI. API. When that evolves. Oh shit. My head hearts.

An API for the Watson computer by IBM
The IBM Watson API, over at Programmable Web.

Hardware becoming software – dimple

Is hardware becoming like software? What is that even? We often hear about hardware becoming more and more like software, meaning open source (-ish), hackable and customisable. Then, of course, focusing too much on hardware might be wrong. Either way.

This thing dimple is quite interesting as it lives somewhere in-between software and hardware and kind of transcends those boundaries. Keeping hardware intact, yet still customising it. By way of software. Hmmm. Something like that.

(if the annoying “support us now” box is in the middle of the screen, hover and click x. But do support.)

Mt. Gox, official systems and a parallel CEO profile

There are some hilarious quotes and secions in this story in Wired, about Mt. Gox and its fall. Simply hilarious, even though there is a serious side to it.

The 28-year-old Karpeles was born in France, but after spending some time in Israel, he settled down in Japan. There he got married, posted cat videos and became a father. In 2011, he acquired the Mt. Gox exchange in from an American entrepreneur named Jed McCaleb.

But soon, McCaleb was getting wires for tens of thousands of dollars and, realizing he was in over his head, he sold the site to Karpeles, an avid programmer, foodie, and bitcoin enthusiast who called himself Magicaltux in online forums.

That’s exactly what I would do with a burgeoning financial empire, sell it to a cat-vid-poster.

No, but seriously. As with everything that has a future affect – before it comes, it comes in numerous versions. This (sort of thing) might do something with how we view the more traditional structures and systems. More positively. Or it might not. Regardless, It has shown that official systems aren’t nessesarily a rule.

The internet of things – industrial internet

The incredibly smart people of BERG hacked a washer and proves a great deal of areas where connectivity help. I mean, the “find repair people” part alone is worth a lot. Some time, after 2 years of really not thinking about it. Postponing rinse takes care of the “shit, sorry I can’t because I’m doing the washing” problem. There are probably not many products that do not benefit from connectivity.

I talked aobut this and that (which is what interesets me most) with a very technically oriented ex-colleague who shared a conversation with interaction designers of a more visual background and nature, and how that hinders the thinking around connected products. “What’s a couch gonna say to me?”. Nada, but tracking the use of it provides input to material choices and manufacturing (something that today is a part of the manufacturing process, but pehaps could be combined and outsourced to “natural use situations”) as well as feedback to healthcare industries benefiting from understanding our couch-potato-behavior.

Cloudwash: the connected washing machine from BERG on Vimeo.

instant-on is the killer TV feature

That’s the central problem plaguing both set top boxes like Roku and Apple TV and content services like Netflix and Amazon Instant Video. Instead of letting you lean back and soak up content, these new challengers require decisions–a careful cost-benefit analysis of thousands of different options. If the traditional TV experience is about letting viewers surf channels, today’s on-demand video is like giving them a speedboat and forcing them choose a destination before they can even get in the water.

From the article “To Grow, Netflix Must Learn From the Quietly Brilliant UI of Regular TVs”, in wired

Well hear hear. This is exactly what Is missing in the creating of the future of TV. From the services currently in the market to the reasoning that goes on in pitch material, and pitches, that I’ve been involved in with regards to positioning TV content providers of today, when developing for tomorrow.

Don’t just stare at new technology and heaps of content at your finger tips. Look at the tired people whose brains stop functioning as their ass lands softly in the couch. The key feature? Bzzz – TV is on and streaming. Streaming something. One more bzzz and there’s something new.

The instant on, under a second, is something “new” TV (content, and the delivery of it) doesn’t manage. This is also why personal, pre-loaded, schedules are important. Not just because you can create your own channels based on favorite content, but because something has to start streaming as you enter couch mode. Human behaviour, not just technology.

ICAs framtida vision en underskattning av tekniken

ICA delar i en film med sig hur de ser framtiden. En framtid de, antar jag, gärna ser och bidrar till. En otroligt krävande, påträngande och underskattande om du frågar mig.

När dina ägg är slut ska du inte bara tala om detta för en liten manick, utan följa upp denna upplysning med en beställning på nya. Kanske även antalet. Det funkar ju, men det är ungefär som de där människorna (ni vet vilka) som talar om att de ska maila, drar iväg mailet, för att 2 minuter senare dyka upp (fysiskt eller i SMS) och tala om att de har mailat. De utnyttjar liksom inte styrkan i tekniken (jo, jag talar om mail som teknik). Predictive technology ska ju ta handlar om just detta, exempelvis google now.

Inköpslistan som skapas av alla dessa krävande inmatningar i ett system (som under de första åren inte kommer vara optimalt utan en orsak till extrem frustration fattar man ju), pushas sedan ut till en stackars familjemedlem (eller slav, det förtäljer inte historien). Snacka om oväsentligt, påträngande strunt. Stackare.

Ska framtiden verkligen behöva vara så krävande?

Nä, så här kommer det inte se ut. Varför? För det första för att det är inte bra, och för det andra för att vi redan har teknik som gör detta på ett bättre sätt.

Egg minder by GE and Quirky
GE har, tillsammans med Quirky, som sysslar med produktutveckling (ofta en “framåtlutad teknisk inställning”) med crowdsourcing som metod, tagit fram en lösning på äggproblemet. Nja, kanske inte ett jätteproblem, men det är ju på detta sättet vi hittar rätt framtida applikationsområden och utföranden – på lite quirky vis… Egg minder hjälper dig ha koll på hur många ägg du har i kylen, samt hur gamla de är. Nästa steg är uppenbart – beställ när antal = x. Du bestämmer så klart, men behöver inte bry dig mer. Inget swipe. Inget inmatande. För varför liksom?

iBeacon är lite på tapeten för tillfället. Det står klart för de flesta att Apple sedan iPhone 4 meckat in iBeacon-teknik i alla lurar. Det är ungefär som jämfört med Bluetooth (det är faktiskt Bluetooth 4.0) vad en full och ansvarslös dörrvakt är jämfört med en nykter – “nej, nej lägg ner ditt leg, jag kommer ihåg dig det är lugnt välkommen in”. Alltså ta tag i en kundvagn och du är registrerad. Lägg ner ditt chip (som du ändå glömde i bilen).

Det är inte framtidens teknik, utan pågående teknikutveckling. Den blir bättre och bättre på vad den är där för; att hjälpa oss. Och i ICAs kontexten handlar detta mycket om att försvinna. Inte att aktivt interagera med oss. Interaktionsdesign och tjänsteutveckling (som den här) kommer, tro det eller ej, handla mindre om att swipa, klicka, peka och dra. Teknik kommer att försvinna, och då gör den ofta ett bra jobb.