But Google does manipulate search results, so ostensibly they should also be put on par with newspapers?
examples:
https://news.ycombinator.com/item?id=27477797
https://news.ycombinator.com/item?id=13274753
edit: Also your comment misleads on what the article reports. Twitter has not lost safe harbor because "they do gatekeeping and have accepted bias", but because they haven't met the staffing requirements imposed by India on large social media companies:
> The move comes as tension escalates between the Indian government and Twitter. Google, Facebook, and several other firms have partially or fully complied with the IT rules, which among other things, requires any significant social media firm (any firm with over 5 million users in India) to appoint a chief compliant officer, a resident grievance officer, and a so-called nodal contact person to address on-ground concerns.
> Twitter has not complied with any of these requirements, the court filing said. Twitter had no comment on Monday’s filing, but has said in the past that it intends to comply with the IT rules.
If you are a global MNC with multi billion dollar valuation, and want to operate in a market that has a billion potential users, it is only fair that the company has a local presence.
What if you're none of those?
Is it still fair to require a local presence for every single site / app / network to have access to a population?
Correct.
Is Google allowed to filter results on any internal criteria they can come up with? It includes all avenues like political, censorship, fake news, low quality, competitor results etc.
If you remove user related data as an input to the search query, the results will be purely based on algorithm efficiency, rather than feeding on the biases of the users.
Such feedback loops, of showing content that is suitable to the user, will lead to information silos and gradual radicalization of the user.
With much respect for those who have to hammer them out: The algorithms are full of bias, they have to be. They are essentially just a pile of hacks. Machines cant understand articles. Even if they could they would have the same problem a human reviewer would have when asked if article A is better than B. You cant switch off your own bias. Even if we could do all that you may end up with a category dominated by the writings of [say] 50-51 year old Caucasian males which forces you to ponder corrections that may or may not be terrible ideas if or if not applied. And then you have to chose which bias to apply and we are back to biased search results.
People are the sum of their experiences, if you remove that nothing is left.
Of course Twitter somehow managed to implement the worst of both worlds, with a sorting algorithm that promotes controversial content (without bias regarding the actual content), but then does biased moderation on that content.
I disagree: the point of the original Pagerank algorithm was its bias. In particular, its use of link weights gave the algorithm a strong bias for (a proxy for) relevance.
An "unbiased" search algorithm would need to return results based on simple criteria, readily explained and presented to the searcher, such as "date first crawled" or "number of keyword appearances."
Please attempt a simple implementation with the methods you describe and see how it performs over even a non trivial amount of data and you will see the difference in results.
And I'm not even considering malicious results gaming the whole system.
I disagree. If anything, I think I give this research more credit, because I appreciate it as imposing a normative view on what makes search results "good."
This is more than just "how do we sort a list containing billions of keys" or "how do we grab data from a distributed database;" it's also "how do we understand relevance or quality based on machine-interpretable signals?" The latter is a judgment call.
Each is ultimately an expression of will, since there is no objective ranking of what makes something relevant or engaging. Everyone will agree on the order of names in a phone book, but nobody will agree on the "relevance" order of search results.
You see this every day in news-adjacent search topics, where popularity, relevance, and factual correctness can all give different "proper" orderings.
Ultimately, people dislike stances that give disagreeable results, but they think benign stances to be invisible. I don't believe a coherent, legal argument can work this way.
I think the essential disagreement isn’t so much about the choice of bias as it is about the exercise of power, specifically on who’s behalf do you exercise power? Do you do it for your users or for the greater good? Or for your customers? Or maybe your investors? Or maybe your ideals?
So Google favoring low latency responses is saying latency (cost) matters when people browsing may not care.
But when those people got their respectable reputation by the same means you could have an entirely circular self-praising reputation machine that seems to work wonderfully but operates without reviewing any work.
At some stage of google the Nr 1 result for both "France" and "Paris" belonged to a blogger from NYC who had a single highly optimized page about France and one about Paris.
Similarly a high page rank page about cars shouldn't be an authority on anything else but if they link to your article about flowers it gains a lot of points. So much it would be attractive to pay them for it.
Back links were fabulous when the internet started then increasingly became a self serving loop.
The only sinister thing here is that, as the article mentions, the Government wants Twitter to reveal the IP addresses of specific users and Twitter is refusing to do so. Also Twitter is marking fake news spread by Government ministers and functionaries as manipulated media.
Another issue that the article doesn't mention is that the Government is pushing users towards Koo, a private alternative to Twitter that is by all accounts extremely aligned with the Government's political and religious views.
It's another step towards replicating the Great Firewall.
Edit: corrected "fake users" to "fake news"
User profiling and data mining, coupled with corporate interest to keep users on the site as long as possible, leads to slow but firm biasing. Search results are not objective, but subjective based on the user history and profile.
On the flip end, it is not a news site, but a massive distributor of news items. So it does act as a newspaper, because people consume large amounts of news from it.
Is it a news site, yes. Does it have mechanisms of editorial policy or checks and balances? Not that I can see.
Is it a search engine. Yes. Is it unbiased as a search engine, no.
But then why is YouTube not being targeted by the Government? One salient difference I see is that Twitter is protecting its users by being identified and punished by the Government for being anti-Government.
It is not a massive distribution network. It is a content heavy site used for sharing videos. The actual sharing happens on other platforms.
And do none of the points in your previous comment matter at all?
So you’re admitting this looks like an act of retribution by the Government then?
Correct. Just a social contagion which profits from sabotaging both news and search.
I’m not sure I understand how the fact that something is enshrined in law means it’s not sinister. In fact, the fact that anti democratic or government strengthening actions are enshrined in law makes it more worrisome rather than less.
This is even more worrying when the “laws kicking in” are laws the government passed unilaterally without discussion a couple of months ago precisely to prevent dissent on social media.
Further, let’s not ignore the elephant in the room. This is a government that has been absolutely pathetic when it comes to competent law making, whether you’re talking about demonetization, GST, or article 370 (which even if you may agree with the action on principle, the execution of which is an absolute sad joke, where the government has unnecessarily imprisoned the people of a state in their own homes for almost 2 years because they had absolutely no plan, and are now looking for a face saving way of backtracking from their actions).
Overall, I see GST as a good thing.
And, for many years, investors from other countries complained about complicated business environment in India and wanted to reduce red tape.
Amalgamating the myriad of tax setups across various states, into one uniform tax code, I consider, is a good thing.
You challenged the notion that GST is a pathetically flawed implementation. So I gave you one. Rest is your call.
GST is a mix of a miserably poor implementation and crass political arm twisting by the party that holds sway in the center to force state governments run by opposing parties to fail.
> Twitter is put on par with newspapers, because they do gatekeeping and have accepted bias.
Twitter does some minimal moderation, and everything& everyone has bias. calling them a newspaper though? no. that's absurd. there's no pretense here that this is truth, that people are reporting. one's entering a bazaar of voices.
I don't get the impulse to be cruel & constraining & mean to these democratic emerging mediums. demanding that they fulfill certain roles we want them to have is unjust, a horrible constraint against creativity & liberty.
Isn't ranking a manipulation? Not sure what you mean by this.
You do not sign T&C with your newspapers do you?
It is extremely sinister that you think a govt has any business curating & culling criticism of its policies or fallacies. That's literally a civil liberty which most politicians in India are rarely aware of unless taught a lesson by the courts.
The dichotomy between neutrality and manipulation is artificial, although I do see the underlying assumption that the output of any algorithm would be "neutral" whereas manual interference with its results would be "manipulation".
In reality, however, the boundary is fluid. Would an algorithm that automatically excludes certain sites be neutral, while excluding the same sites manually is considered manipulation? Is "apply keyword search, then apply blacklist" an algorithm or manual manipulation? On the other side, is an algorithm that automatically penalizes SEO-optimized (spam) sites neutral, even though a search term occurs much more frequently on that given site?
It is absolutely naive to think that companies like twitter don't use that power to advance the political/social agendas their most powerful employees care about and it's even more naive to think that governments will see this power and not want to control it for themselves, but instead let it operate unchecked in their own borders, influencing their own elections. They will either block these companies from operating via firewalls or demand that they use their power to share the political and cultural values they want to push, systematically deplatforming any company that they view as a threat to their agenda. China was ahead of the curve.
Twitter's admins arbitrarily remove content that they disagree with, while allowing content they do agree with, regardless of whether said content is illegal or not.
You take it away and you take away more "free press" from the citizens of India.
So the crux of the whole thing is not the censorship but the entity censoring the content. And obviously the government wants to wield this power.
Why should the attempt to remove some harmful content from their platform create a requirement to be absolutely perfect at doing so? Was there any doubt that twitter had the technical capabilities to remove some tweets?
If you ever help a person, are you then required to help any other person in the same situation?
What's the outcome you want? Do you believe it's possible to moderate billions of tweets without a mistake? If not, what's so great about social media that is overrun with swastikas and porn after the toxicity scared away all normal people? How does that help?
Do twitter employees maybe have the free speech right of not being involved in the distribution of stuff they find horrible? What's worse, your tweets being deleted by twitter or being forced to make statements you know to be wrong and harmful (i. e. coerced speech)?
Do you realise the standard for newspapers in the US is "actual malice", meaning the editors must know something they are publishing is false and harmful, and they must act with the intent to cause some harm. Is that how you'd describe "lost liability protection"?
Twitter can not have the cake and eat it too.
There is nothing sinister just the laws kicking in.
You wouldn't want google manipulating search results and then claim to be neutral.