[vgetr]

I see 2 problems with this. The first is that most of the algorithms they’re using are probably based heavily on machine learning, making them inscrutable not just to the general public but also to any experts they have auditing them. Fact is, Google and Facebook probably don’t know how their tech works half the time.

Second, this sets pretty bad precedent. Asking one of these companies to reveal the algorithms they’re using is tantamount to theft. Why should they hand over their golden goose just because a lot of people want it? What’s to stop the same government to take the technology of a smaller company by force later?

If people are interested in protecting the data Google displays, I think a better solution would be to go after who is leaking the information in the first place (e.g. if Google can crawl HIPAA info from another site, that’s the other site’s fault). If consumers willingly hand over their data to Google, on the other hand, that’s on them.

[jacques_chester]

> Asking one of these companies to reveal the algorithms they’re using is tantamount to theft.

Intellectual property is a legislative creature. There was never a "natural law" conception of it and it largely didn't evolve in caselaw either.

That said, Australia has a constitutional guarantee against compulsory acquisition without payment on "just terms"[0] and which is not for a purpose supported by another Parliamentary power.

Such a case might not succeed (a case by tobacco companies against plain packaging was not successful in arguing that it represented an acquisition of their brands and trademarks[1]), but it would almost certainly reach the High Court.

If their Honours ruled that an algorithm or trained model was property within the meaning of the Constitution, it's likely that compulsory acquisition would require payment of billions, perhaps tens of billions, of dollars. At that point it would be easier to regulate it, which would be unlikely to trigger the guarantee.

I am, of course, not a lawyer. But Constitutional law was one of my favourite subjects before dropping out of law school.

[0] https://en.wikipedia.org/wiki/Section_51(xxxi)_of_the_Consti...

[1] http://eresources.hcourt.gov.au/showCase/2012/HCA/43

[UncleEntity]

> Intellectual property is a legislative creature. There was never a "natural law" conception of it and it largely didn't evolve in caselaw either.

IMO the issue isn't with IP but trade secrets which have a long and well established history within human endeavours. In fact, they are the basis of the patent system which intended to get them "freed" from companies so others could expand on their work.

[jacques_chester]

Trade secrets are often legislative creatures too, especially in the US where it is within the authority of state legislatures to regulate.

In Commonwealth countries it typically arose from equity (through the doctrine of confidence) but has also been subject to tinkering. In Australia it's a mix of confidence and corporations law, the former from equity and the later from §51(xx).

It's not clear to me if a trade secret would be considered property under the just terms clause. I'm sure various law firms will make a mint advising Google, Facebook et al on that question.

[vnchr]

There are other motivations for accountability besides consumer protection. The potential for disinformation is unparalleled. And both have proven in recent years that their judgment is not beyond approach. Trust is no longer an option.

[mc32]

Yeah I think in the long run, some aspects of search are going to be under purview of some regulation.

There is too much potential for SERP manipulation for political ends. Be it regulation, unionization, candidate preference, social issues, etc.

Google and others like it may not overly abuse it now, but there certainly is potential for it.

[protomyth]

The first is that most of the algorithms they’re using are probably based heavily on machine learning, making them inscrutable not just to the general public but also to any experts they have auditing them.

I still think this is going to be a big problem with regards to legislation and safety boards. I'm pretty sure they are not going to accept the black box decision making nature. They see the results of prejudice or unsafe decisions, real or imagined[1], and will probably see lack of ability to explain these actions clearly as hubris on the part of companies. The algorithm did it will not hold as an excuse.

1) iPod shuffle play randomness

[mc32]

With regard to protected data, is there any reason for google to index it? They can filter that out and not index it at all.

[techdragon]

They should hand over whatever they are required to by law if they want to be available in a country that demands they hand things over. This is no different to cooperating with legal warrants and other law enforcement requests.

As to how appropriate the request is, that’s up for debate, but it’s not theft.

[pojzon]

Hand me over any intellectual property you own simply because I want it.. Lets not pretend there isnt a maltitude of pretty dumb laws that were passes simply because lets say "the law makers are not the sharpest tools in the shed"

[qtplatypus]

However intellectual property is the creation of the government.

[pojzon]

Not in my country. Here it is a private property and also a one I can simply "refuse" to give to the goverment.

If they started to extract it by force (the same way some countries steal land) believe me, it will be a desert of dumb people here, because anyone half decent would migrate..

[qtplatypus]

They don’t need to use force to extract it. All they would have to do is pass a law that said that courts would no longer enforce the limited monopoly. If you can’t sue for copyright violation then it would no longer be property.

[chillacy]

> As to how appropriate the request is, that’s up for debate, but it’s not theft.

Yea I guess by definition if the state takes it's not theft. If the state kills it's not murder.