Some thoughts on “be my eyes”
(cross posted from my blog)
In the last few days, the “Be My Eyes” App made quite a splash. And with good reason, as it is a wonderful idea.
The app plans to connect non-sighted people with sighted ones when they are stuck with a certain task. You ask for a pair of eyes, you connect over a smart phone, video the problem you have and get a volunteer human to help you out with a video call. Literally you offer to be the eyes for another person.
This is not that new, for example there were services that allow for annotation of inaccessible web content (WebVisum, IBM's (now defunct) social accessibility project) before. But, be my eyes is very pretty and makes it much easier to take part and help people.
Only for the richer eyes…
Right now the app is iOS only, which is annoying. Whilst the accessibility features of iOS used to be exceptional it seems to be losing in quality with iOS8. Of course, the other issue is the price. Shiny Apple things are expensive, Android devices and computers with built-in cameras less so. The source code of be my eyes is on GitHub which is a great start. We might be able to see versions of it on Android and WebRTC driven versions for the web and mobile soon.
As with any product of this ilk, concerns and criticism happen quickly:
- This may portrait people with disabilities as people who are dependent on others to work. In essence, all you need to do is remove barriers. I know many, very independent blind people and it is depressing how many prejudices are still there that people with disabilities need our help for everything. They don’t. What they need is less people who make assumptions about abilities when building products.
- There is a quality concern here. We assume that people signing up want to help and have good intentions. However, nothing stops trolls from using this either and deliberately giving people wrong advice. There are people who post seizure-inducing GIFs on epilepsy forums, for example. For a sociopath who wants to hurt people this could be “fun” to abuse. Personally, I want to believe that people are better than that, but only one incident where a blind user gets harmed “for the lulz” might be enough to discredit the whole product.
Extending the scope of this app
I don’t see why this app could not become more than it is now. We all could do with a second pair of eyes from time to time. For example:
- to help with some translation,
- to recognise what breed a certain puppy is,
- to help us find inspiration for a painting,
- to learn how to fix a certain appliance in my kitchen without destroying it,
- to have some locals show us which roads are easier to walk,
- to have an expert eye tell me if my makeup looks good and what could be done,
- to get fashion advice on what I could mix and match in my closet to look great.
Some of those have great potential for monetisation, others were done before and died quickly (the local experts one was a product I was involved in at Yahoo called Yocal, which never saw the light of day and could have been foursquare years before foursquare).
Again, this would be nothing new: expert peer to peer systems have come and gone before. When I worked on Yahoo Answers there were discussions to allow for video upload for questions and answers. A prospect that scared the hell out of me seeing that “is my penis big enough” was one of the most asked questions in the Yahoo Answers Men’s health section (and any other, to be fair).
The defunct Google Answers had the idea to pay experts to answer your questions quickly and efficiently. Newer services like LiveNinja and AirPair do this with video chats (and Google, of course may want Hangouts to be a player in that space).
The issues that all of these services face is quality control and safety. Sooner or later any of the original attempts at this failed because of these. Skype services to pay for audio or video advice very quickly became camsex hangouts or phonesex alternatives. This even happens in the offline world — my sister used to run a call centre and they found out that one of their employees offered her phonesex services to eligible men on the line. Yikes.
Another issue is retain-ability and re-use. It is not fun to try to find a certain part of a video without a timed transcript. This can be automated to a degree — YouTube's automatic subtitling is a good start — but that brings up the question who else reads the private coaching session you had?
Can this be the start or will hype kill it again?
If anything, the user interface and interaction pattern of Be my Eyes is excellent, and the availability of video phones and chat abilities like WebRTC make it possible to have more of these services soon.
In the coding world, real live interaction is simple these days. JSFiddle's collaboration button allows you to code together, JSBin allows people to watch you while you code and Mozilla’s together.js allows you to turn any web page into a live audio and video chat with multiple cursors.
We use Google Docs collaboratively, we probably have some live chat going with our colleagues. The technology is there. Firefox now has a built-in peer to peer chat system called Hello. Wouldn’t it be cool to have an API for that to embed it in your products?
The thing that might kill those is hype and inflated demands. Yahoo Answers was an excellent idea to let the human voice and communication patterns prevail over algorithmic results. It failed when all that it was measured against was the amount of users and interactions in the database. This is when the quality was thrown out the window and “How is babby formed” got through without a blip on the QA radar.
Let’s hope that Be my Eyes will survive the first spike of attention and get some support of people who are OK with a small amount of users who thoroughly want to help each other. I’d like to see that.