Hacker News new | ask | show | jobs
by noad 2261 days ago
You're forgetting the main benefit for google, which is getting humans to train all their vision models for free. At one point they were just forcing X% of clicks to fill out a captcha regardless of origin or identity just to get more data.

I for one am getting quite tired of trillion dollar corporations getting things for free out of me. Hard pass.

6 comments

> You're forgetting the main benefit for google, which is getting humans to train all their vision models for free.

Is this still true? I keep seeing the same type of images for years and there might be 7 or 8 different categories but that's it. To me reCaptcha looks like a service well in its maintenance phase. If it was actually in use for training purposes you might expect images to match a wider range of tasks.

I could swear I've seen challenges with night scenes (low light conditions) in the wild. Those were definitely not present earlier.
I've lost track of how many times I've had to read house numbers from Google Street View...
I haven't gotten one of those in years. These days it's just picking out buses, cars, traffic signals, and sometimes motorcycles. Maybe once in a while it'll ask for storefronts.
Most of mine lately have been traffic features also. This is a little tricky in some cases, e.g. with crossings, as it sometimes gives me things that I don't think are crossings but it insists I select, perhaps they are in the US, or the perspective is weird, or someone else has told it that a series of white squares is a crossing and it requires me to agree.
Like trees, bridges, fire hydrants, cars, buses, house numbers, etc?
Except in this wonderful new world, you don't get the choice to "hard pass". As someone whose ISP has too few public IP addresses, I see Cloudflare's "one more step" pages at least several times a month. It's terrifying to realize just how much of the internet is behind that thing right now.
This really shows how popular perceptions of Google have changed for the worse over the years. I remember when RECAPTCHA was first launched, everyone knew right away that it was just helping Google train their vision models, but at the time we all thought it was cool, like "Wow, I'm helping the cause of AI research at the same time as stopping spam". But now it just pisses everyone off.

Hell, for a little while Google had a game (can't remember the name of it) which was labeling images with another person to get points and people loved it.

At least the original reCAPTCHA was used for OCR'ing public domain books. Even if it had the effect of training Google's OCR tech, it was at least making literature searchable and indexable for the public good. Modern reCAPTCHA is nothing more than training for Google Maps and, seemingly, self-driving cars, both of which are commercialized.
> But now it just pisses everyone off.

Though we're still just talking about a few HNers here who complain about doing "free work for Google", not the broad population.

I really don't think the challenges we're giving at still hard for computers.. a lot of these are super simple.. google would've cracked many of the driving ones years ago
If that was still the main benefit for them, they wouldn’t be planning to start charging for it, because that would—and, as this article shows, has—cut off much of that data flow, as reCAPTCHA clients abandon the service for another one that isn’t charging them.
Did you even RTFA and look at hCAPTCHA? hCAPTCHA couldn't be more grossly focused on neural-net training. Hell, one challenge asks you to draw a bounding box and another is a classification tagging.
There was no argument being made for HCAPTCHA in the post to which you replied. So, yeah, everything you mentioned is indeed gross, including Google's behavior.
The parent post was edited.