My friend Ian Forrester asked me for my thoughts a few months ago about the Internet Health Report that Mozilla published earlier this year. If you haven’t read it yet, I strongly suggest you scan the site. It’s a great document that covers some important issues with the Internet — what makes it strong, and what work is needed.
I especially like the 5 pillars of a healthy Internet: open innovation, digital inclusion, decentralization, privacy and security, and Web literacy. There are great examples in each area on the health report covering some of the historically important issues that organizations like Mozilla and its allies have addressed.
But I have some issues that are important to me personally that I felt were not called out in this list. They’re mostly forward-looking; paying attention to parts of the Internet that are just emerging.
- AI. This is what I care about most. Current AI techniques require having lots of data, which limits the number of participants. It’s mostly governments and big commercial orgs creating and deploying AI today. Individuals, ad-hoc groups and non-profits hardly use it at all. That’s going to cause quite a skew over the next decade.
- VR. VR is sliding very much into closed systems like Steam or the Google Play Store. There are not open VR explorer systems in wide use. WebVR is a good first step, but we need to see more deployment and usage.
- Voice interfaces. Siri and Google Assistant are hugely centralized system; there are only a few other players. They are not open systems; it’s hard for developers to add new behaviours to Siri, for example. And it’s almost impossible for end users to correct voice interfaces (“No, that’s not ian’s email address”) or do end-user programming (“tell me any time ian sends me an email about mozfest”). The fact that most speech-to-text systems are cloud-based (everything you say gets sent to the cloud for recognition) is a potential nightmare for privacy.
- Touch-based software creation. Almost every interface in computing has changed radically since the 1950s with the exception of software creation. We still use an antiquated model of creating text files and running them through a compiler or interpreter. But most computer users today use touch-screen devices. Why don’t we have more touch-based software creation tools?
- Dating! I realize it seems trivial to some people, but romance and sexuality are a huge part of human existence. Many major dating sites are owned by a single company (IAC). The network effect make decentralized dating very hard to pull off. It’s an area that requires privacy and gradual disclosure. Open dating systems would be fascinating — posting one or more profiles on the open web in a way that preserves your privacy but allows gradual disclosure and connection.
I think there’s a lot more that needs to be addressed. I’m facilitating sessions on democratizing AI and on open dating as well as giving an update on the ActivityPub network at Mozfest 2017 this weekend.
Evan Prodromou's Blog 