Posts Tagged ‘candid a.i’


I have previously written a piece on Candid and its A.I as well as mentioning Google’s Jigsaw and the possible dangers it posed to freedom of expression on the internet. Initially, I believed Candid’s A.I to be flawed but relatively harmless in the grand scheme of things. I should never have been so naive.

From the short time, I used it. I quickly found the app to be an unorganised mess. There was no real discussion, like people were being offensive for the sake of it. There wasn’t a lot of productive discussion in any of the groups and unlike Minds, you’re limited by how much reach you have. You also risk being damned to the random group if your post is considered to be offensive. Or worse have the post removed without so much as a notification.

It’s already been mentioned how Harmful Opinions video criticising Candid could not be posted but as time has gone by the CEO Bindu Reddy and those under her employment have engaged in a witch-hunt of anyone critical of the app. Claiming that there will be legal consequences and that she has a case for libel.

This all appears to be coming off the back of an Encyclopaedia Dramatica article that goes into some depth on what lies behind the code of Candid. If you are interested in reading it, then here is the archived page. However, I do suggest reading it outside of work since ED has a lot of NSFW content dotted around its pages.

The first to observe is that Candid is to some extent recording the details of those who chose to connect their facebook accounts (really defeating the point of anonymity) although you can skip this. The button to do so is relatively obscure. The real concern is that Candid is data-mining its users using an app called Kochava. A quote from MobyAffiliates in the ED article describes Kochava as:

‘Kochava is a mobile app marketing tracker with a unique approach, it looks at all device identifiers as equal and as such is able to match the identifiers of different publishers to provide effective analysis and reporting to advertisers. In addition to this, Kochava also automatically engages a device fingerprinting system, using a number of algorithms incorporating carrier and geo-location to match clicks to installs with an accuracy rate of 85%. Offering deep level integration support, Kochava supports server-to-server integration as well as an SDK for Android and iOS. Match reporting for each attribution includes how (device, hash types etc) and Cohort analysis is offered for ROI overlay as well as optimisation according to various campaign metrics (clicks, installs, post-install revenue etc).’

The ultimate point is that Kochava is using your information to feed you ads. For a service built on allowing users to be anonymous. It certainly seems to be doing the exact opposite of this. The ED elaborates on how connecting Facebook allows Candid access to your feed, your likes, your app invites and your messages. This data-mining extends to knowing the model of your phone and even your cellphone number if its present on facebook. But even without connecting to Facebook, more code reveals that your location is being tracked. It’s also sorting your apps into lists of whether based on quality.

The ED article is ultimately damning of what has gone into this apps programming. However, it gets much worse, after Mark Kern spoke negatively about the app. He was forced into silence by the actions of those at Candid since they had begun digging into his past. And when Harmful wanted to talk to Reddy on stream she made the request that ‘comments be disabled’. For someone who has created an app centred on free speech this is immensely hypocritical.

This situation in all fairness has been overblown and Reddy fanning these flames has merely caused the Streisand Effect. If this escalates any further into legal implications, I honestly wouldn’t be surprised. Reddy and those working for her seem to not realise that criticism is allowed in today’s society. For those under her to actively seek to ruin the lives of their critics is abhorrent by itself. But in the age of the social media mob, I’m just disappointed. I will note that she has since apologised for these actions but they never should have been undertaken in the first place. If you want to hear more about all this, Reddy did go on a stream with Bearing, and Harmful posted his own response to that.

If any more developments occur, I will probably write a part 3. If you haven’t read part one, you can find it here. It mentions Jigsaw google’s AI. That has already policed a few comment sections and is going to save humanity from itself.
God help us all…