Android developer verification: Balancing openness and choice with safety

News and insights on the Android platform, developer tools, and events.

Android Developers Blog

At this point I'm convinced that there's something deeply wrong with how our society treats technology.

Ruining Android for everyone to try to maybe help some rather technologically-hopeless groups of people is the wrong solution. It's unsustainable in the long run. Also, the last thing this world needs right now is even more centralization of power. Especially around yet another US company.

People who are unwilling to figure out the risks just should not use smartphones and the internet. They should not use internet banking. They should probably not have a bank account at all and just stick to cash. And the society should be able to accommodate such people — which is not that hard, really. Just roll back some of the so-called innovations that happened over the last 15 years. Whether someone uses technology, and how much they do, should be a choice, not a burden.

> Ruining Android for everyone to try to maybe help some rather technologically-hopeless groups of people is the wrong solution.

This isn't about how skilled a person is, it is about tackling social engineering. The article gave the example of someone posing as a relative, it could also be a blackmail scheme, but it could also be the carefully planned takeover of a respected open source project (ahem, xz).

What I am saying is this sort of crime affect anyone. We simply see more of it among the vulnerable because they are the low hanging fruit. Raising the bar will only change who is vulnerable. Society is simply too invested in technology to dissuade criminals. Which is why I don't think this will work, and why I think going nuclear on truly independent developers is going to do more damage than good.

There's quite a gap between this sort of opportunistic scamming that's happening all over the world and targeted multi-year campaigns that probably require the resources of a nation state.

> targeted multi-year campaigns that probably require the resources of a nation state

Ha ha ha, "resources of a nation state"! One could run phishing campaigns at scale over many years without breaking the bank. This was true before LLMs, it's probably even cheaper now.

Sorry, I keep forgetting that LLMs are a thing. But I disagree because many people, especially tech-savvy people, can't possibly trust any communication that has the hallmarks of slop.
At this point it’s naive and perhaps a bit dangerous to assume that any of us can differentiate LLM from non-LLM text. I see less and less recognizable “slop” as time goes on, but I doubt the amount of content being generated has gone down.

especially tech-savvy people, can't possibly trust any communication that has the hallmarks of slop.

And yet, people on HN respond to bots all the time.