Hacker News Reader

Comment by rsa25519 | Hacker News Reader

> I don't think Google actually made either of these claims in their letter.

Agreed. And because this ACCC response doesn't link to the letter, I think these rebuttals to non-existent claims are deceptively misleading

View on HN

The google letter makes non-existent claims that aren't part of the proposed law. Like that youtube and google will somehow magically offer worse results. They can offer the exact same results they do now - they might just have to pay for it. "If we had to pay more for news content, we would probably just cut it out because we don't want to pay - thus leading to worse results" doesn't quite have the same ring, does it :-)

As such the ACCC response is also full of bullshit. This is how politics go, and the sad part is that most people seem to simply read one side of the argument and go ranting off angrily based on the bullshit claims, rather than investigating the truth behind both sides

> they might just have to pay for it. "If we had to pay more for news content, we would probably just cut it out because we don't want to pay - thus leading to worse results" doesn't quite have the same ring, does it :-)

The amount they'd have to pay for it seems to be dramatically more than what they make for it, so "if this law passes, we will stop serving these results" seems to be a reasonable reaction.

> "if this law passes, we will stop serving these results" seems to be a reasonable reaction.

If that's Google's reaction then they should say so outright, not this weasel wordy letter.

Being required to inform news media about changes to result ranking algorithms in advance and advise them on how to avoid being impacted also gives news media a unique ability to game search results and takes away Google’s ability to prevent them from gaming results. I would certainly expect this to also seriously degrade the quality of search results. Not only will results become a pile of worthless SEO spam by news companies, Google will also have to use a much more simple ranking algorithm free of both ML techniques that can’t be explained to news companies and advanced algorithm techniques that are valuable trade secrets that news companies would obviously leak with their many employees, poor security, and adversarial motives.

I would just expect Google to use obsolete or different algorithms for search results for Australia.

The "worse results" probably stem from the part of the law about changes in pagerank algorithms: If search engines are required to notify about changes in algorithm and provide mitigations for those changes, the easiest way to provide this would be: Keep the algorithm as it is, never change it again (for Australia) and accept the degradation that will result from that.

Also, what does this section even mean in the context of ML models? How do you communicate that you trained an interest-estimation model on more data and will now be using the upgraded version? I’m sure Google uses many such models for producing personalized results, and probably several are updated in production every day. Does Google have to share the inner workings of their search stack with these news companies? Do they have to share the models? Does Google have to explain to these news companies how the newly trained model might treat their content differently than the previous iteration? Any of these things I expect would cause Google to throw up their hands and abandon the country. Exposing a live feed of all their trade secrets to the employees of random news companies is obviously not going to happen, and explaining why ML models do what they do is an active research area.

These are excellent points about the ML aspect that I haven't seen mentioned so far.

Not only that, it would allow SEO companies to sign up as "news organizations" and then use the knowledge gained about the ranking algorithm to fill everyone's search results with SEO spam even more than it already is.

The legislation would prevent Google from removing the news content from search results.

Isn't that a basic tactic in political communications: when asked a question provide the answer you want to give not the answer to the question you were asked?

Similar to "strawman" in online forums, right

More "message discipline" I suspect:

https://en.wikipedia.org/wiki/Message_discipline

Interesting. I wonder who coined that phrase? It's a curiously positive sounding phrase for something that's actually quite nefarious.

The discipline, applied to its own name. :-)

arethusa, thanks, interesting.

I guess this is something taught in paid classes to business leaders and politicians, so they want a positively sounding name.

Also, except for when the person has evil / corrupted intentions, then, is this technique neutral or sometimes even good

Can at least save time via fever of topic discussions

From google’s open letter

> Under this law, Google has to tell news media businesses “how they can gain access” to data about your use of our products. There’s no way of knowing if any data handed over would be protected, or how it might be used by news media businesses.

That seems to pretty clearly suggest Google will be forced to hand over personal data.

It's deceptive. They aren't saying they would hand over any data.

That seems to be pretty clearly what the law says they will have to do. Do you have a different interpretation?

It’s deceptive to directly quote them?

Reply on news.ycombinator.com