Podcast Addict app to be reinstated, Google apologises
9to5google.com
9to5google.com
Treating the developer with suspicion, after they clumsily and callously trampled over him. Insult to injury.
That spokesman needs to be fired, and we need to stay angry until Google gives us developers a real pathway of appeal/communication concerning our apps being taken down, so we're not talking to a cold machine when our livelihood is suddenly yanked out from under us. We shouldn't have to make a public fuss to get our apps reinstated.
Typical Google at this point. Maybe I don't understand the scale, but wouldn't it make sense to have _some_ kind of human in the loop? Given some criteria, perhaps an app has existed for >5 years, meets some threshold number of users, etc. maybe it's worth getting a human opinion before doing something as drastic as banning.
But the evaluation statistics tend to be averages, rates of false positives/negatives, and other macro-level things that often don't distinguish how some items (apps, in this case) are more important (popular or economically valuable) than others.
I've been in meetings all the time where this has happened -- the data scientist without a good handle on the business presents the statistics, and the product manager who knows the business trusts the data scientist. Hopefully someone with common sense will pipe up "but wait... what if we accidentally take down a really popular app?" But because it's 1 thing of 20 that are on the meeting's agenda, people miss it and it just gets approved.
And the further problem is that there's no standard process (that I'm aware of) for determining whether a data science model meets the actual business requirements, except a lot of trial and error that often involves actual business failures. It's pretty frustrating, really.
Because when it used to just be hard-coded rules, someone was responsible for qualitatively evaluating those rules, and common sense prevailed a bit more. But when data science models are often more of a black box, you evaluate with statistics, and what does it mean to then bring common sense in back on top of that?
It's brutal and harsh but kind of ..understandable and justified when looking from their point of view, there are plenty of tiny-teeny fish in the ocean and if some die? well, it happens.
There is also another lesson and message - think twice before you completely and without any fallback embed yourself into wonderfully crafted and cushy ecosystem offered by one of giants. Maybe web version of your app[0][1], accessible by everyone and outside of garden could be useful after all?
[0] https://i.imgur.com/ovqMkU5.png [1] https://i.imgur.com/wHOA1DP.png
How do we correct our mistake ? I search on duckduckgo from firefox and switch to losers when something gets too big, but I fear the effect is limited. And switching to LBRY from Youtube is... difficult :D