John’s AI Reading Links

As we get ready to fully launch our newest conference, Web Directions AI, I’ve pulled together some particularly interesting perspectives on AI for your reading pleasure.

Watch out for an email today with all the details of the one-​day conference in Sydney on 28 September.

Cheating at AI

Hopefully, it’s clear by now that I think an area anyone — whether  more on the design, engineering or business sides of the Web/​Digital/​Technology area — should be focusing on involves what we broadly call AI.

So much so that we’re about to launch our brand new AI focused conference, taking place in Sydney in late September.

If you want some quick overviews of the key ideas, Stefan Kojouharov has assembled a list of “cheatsheets” on machine learning, neural networks and more related topics.

It’s 1996 all over again

Ever fewer of us in the field remember the website designs of 1996: the year “Killer Web Sites” dominated the web design conversation and the year of the first US Presidential election to take place in the consumer web era. For increasingly many, this is a lifetime ago.

Dole/Kemp 96 website - no AI

Putting today’s sophisticated interactive web experiences up against this makes you wonder how even we got from there to here.

But all technologies take small incremental steps from the realm of the early adopter to the mainstream, and today’s hotness, chatbots (I prefer the broader idea of conversational interfaces), are very much in their infancy.

Vittorio Banfi makes the argument that Chatbot design today is like web design in 1996, and who wouldn’t want the opportunity to get in the time machine back to the start of web design and help shape that field? Well, maybe this is your chance.

Machines versus abuse

The sheer scale of modern social platforms like Facebook, Instagram and Twitter, with hundreds of millions and even billions of active users a month, makes any sort of human powered curation and moderation laughably impossible.

Perhaps this is why these platforms seem so often to be associated with negative human behaviours, from fake news to bullying and abuse. These behaviours can be difficult to detect and respond to. There’s also the challenge of company valuations being a function of active user numbers, driving short term disincentives to removing users, even fake ones, from the network.

Recently, Instagram has begun turning the potential of parent company Facebook’s machine learning engine DeepText to the challenge of bullying and abusive behaviour on its platform.

More at Web Directions AI

If you’re keen to learn more, keep an eye on Web Directions AI  taking place in Sydney at the end of September, with program announcements starting next week.

From design to business and technology, if you have even the vaguest inkling that this stuff is going to be important in what you design and deliver, don’t miss AI!

Comments are closed.