Live data from Hacker News

Google and Paint.NET need to stop misleading users (2013)

jasonlefkowitz.net

51–60 of 109 posts

Re: Google and Paint.NET need to stop misleading users (2013)

#51
post #44

Earlier quoted context omitted.

Please explain how, since it’s easy.

I'll take a crack at an explanation: Google has machine learning technology that automatically reads street address numbers on houses, regardless of angle, color, size, focus, or typeface. Using the same type of algorithm to answer "does this ad have anything that looks like a button in it?" is relatively simple by at least two orders of magnitude. So rather than it being "easy" in some absolute sense, what's meant h…

Lots of things have button and aren't maliciously masquerading as download links. Furthermore, if they start doing that then ad producers will start changing their ad images to evade the algorithm. It's much harder to use ML for a problem when you have a malicious opponent actively working against it.

Re: Google and Paint.NET need to stop misleading users (2013)

#52
post #30

https://ninite.com/ Accept no substitutes when installing on Windows.

Looks nice, but I would always prefer Chocolatey ( https://chocolatey.org/packages ). Reason: More (and easier to customize) packages / applications to get.

Ninite has the advantage of being self-contained in a single executable, though. Chocolatey looks like something you have to install, which makes it poorly suited for quickly installing something on a machine that's not yours.

Re: Google and Paint.NET need to stop misleading users (2013)

#53

Sounds like a personal problem to me... This is a little shady, but all the units are clearly labeled as ads. I have a hard time getting as worked up over this as the author does. Paint.NET is a free product, what do you expect?

You could've at least visited the website. They are NOT clearly labeled as ads. https://www.getpaint.net/index.html

wcummings might mean the tiny grey-and-blue icons in the top right of each unit.

If so, I disagree. Those icons are not visible enough, and are hard for many users to click without accidentally clicking the ad instead.

Re: Google and Paint.NET need to stop misleading users (2013)

#54

And as unpopular as the idea is with geeks, I think this illustrates the need for a "walled gardened" for the majority of users where each app is sandboxxed. Something like this couldn't happen on iOS, ChromeOS or Windows 10s. Most non developers would be served perfectly fine with a locked down OS.

What it illustrates the need for is understanding why you're receiving things (which is a pretty vital outside of software, too). If you pay for your product then you won't see phishing on the download page. If it's a subscription fee, then they're going to make it difficult to abandon or switch away from. And if it's "free", then who knows what's going on - maybe the product will wantonly encourage spending money (advertising), or manipulate you towards a paid service. Or maybe the secret motives are innocuous, like the dev's resume or (very rarely) the common good, but you need to understand that before you know what's coming.

Re: Google and Paint.NET need to stop misleading users (2013)

#55
post #51

Earlier quoted context omitted.

I'll take a crack at an explanation: Google has machine learning technology that automatically reads street address numbers on houses, regardless of angle, color, size, focus, or typeface. Using the same type of algorithm to answer "does this ad have anything that looks like a button in it?" is relatively simple by at least two orders of magnitude. So rather than it being "easy" in some absolute sense, what's meant h…

Lots of things have button and aren't maliciously masquerading as download links. Furthermore, if they start doing that then ad producers will start changing their ad images to evade the algorithm. It's much harder to use ML for a problem when you have a malicious opponent actively working against it.

There is a human behind every single creative shown on Google AdSense. There are even rules around these creatives (need to have distinct border and need to say the name of the advertised product).

Re: Google and Paint.NET need to stop misleading users (2013)

#56
post #51

Earlier quoted context omitted.

I'll take a crack at an explanation: Google has machine learning technology that automatically reads street address numbers on houses, regardless of angle, color, size, focus, or typeface. Using the same type of algorithm to answer "does this ad have anything that looks like a button in it?" is relatively simple by at least two orders of magnitude. So rather than it being "easy" in some absolute sense, what's meant h…

Lots of things have button and aren't maliciously masquerading as download links. Furthermore, if they start doing that then ad producers will start changing their ad images to evade the algorithm. It's much harder to use ML for a problem when you have a malicious opponent actively working against it.

It should be as simple as:

- Is this a download page?

- Does this ad contain a button-like image with the word "Download" on it?

Presumably the people creating these ads are specifically targeting pages with the word "download" in to create this confusion, it doesn't seem like rocket science for Google to apply the exact same rules.

Re: Google and Paint.NET need to stop misleading users (2013)

#57
post #51

Earlier quoted context omitted.

I'll take a crack at an explanation: Google has machine learning technology that automatically reads street address numbers on houses, regardless of angle, color, size, focus, or typeface. Using the same type of algorithm to answer "does this ad have anything that looks like a button in it?" is relatively simple by at least two orders of magnitude. So rather than it being "easy" in some absolute sense, what's meant h…

Lots of things have button and aren't maliciously masquerading as download links. Furthermore, if they start doing that then ad producers will start changing their ad images to evade the algorithm. It's much harder to use ML for a problem when you have a malicious opponent actively working against it.

ML could be used as a tool to empower the reviewers to to review the most suspect ads first.

Re: Google and Paint.NET need to stop misleading users (2013)

#58
post #7

This is trivially solved with an ad blocker. Frankly, I do not see the problem. Why should you rely on Google/Paint.NET/state legislation when the solution is literally a few button clicks away on both chrome and firefox? All browsers should come with ad blockers by default. Hell, the OS should come with blocking hosts.txt. That would make the digital world a whole lot safer. And before you say the internet will die,…

I don't see any problem too. Now they download something that is not paint.net , next they download and run malware. In this case the cost of mistake is cheap. Let them learn from their mistakes. What we need is to educate people how to download and run what they really need, not blindly clicking the biggest banner on the page.

> In this case the cost of mistake is cheap. Let them learn from their mistakes.

All well and good until they unwittingly become part of a botnet and start attacking the rest of us...

Re: Google and Paint.NET need to stop misleading users (2013)

#59

The terrible experience of pulling down an installer for Paint.net was one of the final straws that pushed me to move over to using chocolatey for windows workstation setup. > choco install paint.net Unfortunately, while that allows me to never see fake download-button ads when pulling down software for my machine, one place it's still a nightmare in is the world of getting minecraft mods installed on a kid's compute…

Yep, the only time I've ever managed to install malware was trying to put a very ordinary Minecraft mod on for my kid. How to feel stupid in one easy move.

Re: Google and Paint.NET need to stop misleading users (2013)

#60
post #45
post #31

You can report these bad ads by clicking the small X in the corner. Please do that if you ever see any of those, and they'll be replaced by other kinds of ads (and hopefully over time they'll get enough reports to be removed.)

The better solution is installing a capable adblocker.

That works for me, but then leads to me telling a friend/relative to get the software and them managing to screw up their machines by doing what they think I told them to do.
Post reply on HN