Flying cars and more: 6 Cool Ideas from Web Summit 2017
Over the final 2 days of Web Summit, I attended talks at the binate.i(o) conference, which is a data conference, FullSTK, the developer’s conference, as well as one or two talks at HealthConf and some of the Centre Stage talks which tended to focus mainly on AI.
I have to admit I was rather disappointed with the FullSTK conference as there was a distinct lack of what you would define as actual code. However, I did still learn a ton of interesting stuff — in fact, I learned so much about data and security that I think this subject deserves a post of its own which you can find here. As for the rest, here’s what I found!
1. Flying cars will very shortly be a thing!
Uber is set to launch flying cars in LA in 2020 — only 1 year after this was predicted in Blade Runner, as Uber CEO Jeff Holden was stoked to point out. They will be piloted (though a transition to driverless flying vehicles is certainly on the cards), emissions-free and travel at speeds of 150–200mph, initially between specially designated ‘skyports’. That’s all great but while I was actually at the conference, mere hours after being dazzled by all of this, I received an email saying that Uber drivers would be striking in Portugal that day. Oh, the irony…I can’t help feeling that they need to start paying rather more attention to their drivers and regulations if they want to achieve all that they have in mind. That being said, Holden’s vision is that in the future, only hobbyists will own cars as it will be both cheaper and more convenient to use Uber self-driving vehicles. I don’t have a crystal balls, so I guess we will just have to wait and see.
2. Graph databases can find connections that no other technology or people can see
I also attended a talk on graph databases by Emil Eifrem from Neo4j, who explained that with nodes, relationships and properties, you can model everything, and that what this means is that big data can now by used to make connections. These would otherwise go undetected in the classic database store, as when you query the data, instead of just getting the data you asked for back, you get back information on the context of connections that are being made. He likened a graph database to the human brain which is also joined through many connections. This has amazing real world applications, such as finding causative factors of cancer and the most effective treatment methods for patients.
3. Ways we can make AI usable for everyone
Danny Lange from Unity Technologies told us that in order to make make this happen, machine learning needs to be:
· Easy to use
· Pervasive and versatile
He also came back to this idea that machine learning systems are trained with human bias and imperfect data, i.e. you get out more of what you put in. And then we saw a rather cool video about how a robot learned to cross the road by itself, by training the software initially in a virtual environment where it can’t be squished by an oncoming lorry before putting it in a robot that went out into the big wide world. It was sort of sweet actually — reminded me of Wall-E.
4. Arduino is getting easier
Massimo Banzi, Co-Founder of Arduino, told us that he wants to enable people who perhaps have less technical knowledge to be able to get to grips with the Internet of Things and create cool stuff . I think this is great. The cynical side of me also says that this is a great business model, because the more people that can use it, the more people that can buy it. Nonetheless, this is sparking some surprising innovations — my favourite example is a chair that tweets when you fart. Perhaps not the most useful tool in all fairness, on the other hand, it makes me laugh to know that this exists in the world somewhere and I’m all for innocent ways of bringing people joy.
5. You may one day be prescribed a Fitbit by your doctor
It was also great to hear what Fitbit has in store. They plan to link Fitbit to your GP so that exercise plans can be administered and data can be compiled, which can further knowledge of conditions on both a personal and global level, therefore better informing doctors about which treatment to give. The aim is to lower overall treatments costs, as drugs can be targeted to individuals much more accurately.
6. Mozilla have a new browser — and it’s better than Chrome! (at some things)
Finally, Mozilla showcased their new browser, Firefox Quantum! And talked us through some of the issues they experienced while updating it and how they managed to get it up to speed and be a competitor to Chrome once again.
Quantum is also a multi-process browser, using a new language, Rust, instead of C++, which they said was fast but buggy and therefore expensive to maintain. They really loved Rust, actually, and told us we all need to go out and give it a go. I just might. If I’m not too busy attending AI and machine learning meetups after this week 😉 Mozilla also uses open source libraries, and they succeeded in mobilizing the developers who worked on these to reduce the number of bugs in their code, which is pretty impressive in my opinion. Though I guess as a developer, it would also be pretty cool to be part of bringing a new browser that could rival Chrome to birth. At the end of the talk, they showed us comparison videos of Chrome versus Quantum performing various tasks and for some things, most surprisingly Google search, Quantum was actually faster. However, this was by no means always the case, and it seemed to be a fairly even 50–50 split of which browser was the quickest, depending on the sites visited.