More evidence that Google manipulates results to promote its politics, which they have made clear are opposed to Trump and his voters. They always assure us that what they talk about never was implemented. Of course they can offer no proof of that. Also, they obviously don’t care about employees talking about destroying their objectivity if these people are still employed.
If they did would people be able to understand and interpret such evidence? It’s a mixture of code, trained AIs, and custom hardware. I don’t think most people have the digital literacy to even begin to mentally process it.
Certainly this is something the community of well educated computer professionals could determine. That’d be good enough for me.
I’m sure it’s more trouble than it’s worth though. To be honest.
Not really. Anyone who has worked in software knows that there is a huge difference between discussing something in email and actually putting it into practice. In fact, at least for my team, email is probably the worst place to open up an idea if you want it implemented.
And of course you have no proof that it was implemented.
Google doesn’t strike me as a company that plays mommy to their employees. At least in my experience at a similar company, the culture tends to be more laid back regarding what you can say and do at work.
I think looking at the technology isn’t the only way to prove Google implemented this. Sure, source code could leak and potentially indicate that they did, or a bot could run numerous tests and try to reverse engineer that part of the algorithm, but there are other ways.
For instance, I’d imagine that Google has ways of tracking who worked on what features and when they were implemented, similar to how Microsoft has
VSTS Azure DevOps. That would be pretty easy to understand for non-tech people, and I’d imagine whoever had access to these emails to leak them has access to all of that.
Also, if they use Git (which is very likely), there’s also Git commits and PR descriptions, which would also be easy. Again, if the source of this leak had access to these emails, they probably have access to the Git history as well.
Ultimately, in terms of stuff that could have been leaked to prove a feature was implemented, emails are about the worst possible thing to leak. Sure, it could have been concern over leaking business-critical stuff like source code, task assignments, and Git history, but if this were implemented, then I’m sure these emails are considered business critical and just as job-ending to leak as any of that stuff.
Definitely likely. I pull from one of the Chromium repositories and Tensorflow repositories which they publish via Git.
That’s part of why I think they use Git internally.
Git is also just industry standard. Even Microsoft, who had their own source control system in TFS and is notorious for sticking to their own tech, eventually moved to Git.
No I don’t nor could I. But we do have proof that employees conspired to do this. They thought it both possible and a good idea.
If I found out my bank had employees (who are still employed) who conspired to take a few dollars out of everyone’s accounts I’d switch banks. I wouldn’t really worry about claims that they didn’t implement it.
Often times at large software companies access to code is very controlled. It isn’t a safe assumption that someone with access to some emails has access to this code. Even if they did it wouldn’t matter. Google and their apologists would just say you can’t prove that is the actual code running, and no one could.
You could if something better than emails leaked to the press.
From a technical perspective, it is obviously possible. That’s never really been a question. Sure, it might be difficult to implement without breaking a lot of stuff, but again, not impossible.
However, knowing that something is technically possible and wanting to do it are, again, far from doing it when it comes to software development. There’s planning, code writing, code testing, and code review that it has to go through, to oversimplify what actually goes on.
Without knowing much about banking systems, I’m not sure that they are equivalent.
What we’re talking about here is an algorithm made in reaction to a major controversy. Considering the simplified process I mentioned above, such algorithms are almost never make it through because:
- They are reactionary to a problem whose outrage is likely going to fizzle out quickly. Most companies are thinking more long-term and won’t tolerate potentially long-lasting damage from code changes that are that short-sighted.
- It’ll have to go through a lot of review, especially if it is as controversial as this. By the time it time it even reaches that stage, the outrage may have died down, and it’ll get a lot of “Why bother?” comments.
In the banking system scenario you gave, their obvious path is circumnavigation of the systems in place to protect customers. In this case, we have little evidence of any attempting circumnavigation, and the systems in place would likely catch them.
I’m not 100% sure, but from what I’ve read on Google, their repos are open to all employees regardless of project. That’s one of the things that make them unique compared to Microsoft or Apple.
But even assuming a more locked-down company, things tend to operate on a group policy, and that group policy is likely associated with the team and/or product. Considering that these emails were about algorithmic changes, we can almost certainly assume that those receiving the emails were part of groups capable of making changes, and those groups would have access to the repo.
So whether we’re talking about Google, Microsoft, or Apple, it seems reasonable to assume that the person who leaked this email could leak Git history if something incriminating were contained there.
That is in different territory. That’s fiduciary theft and since it’s a bank it is possibly a federal crime. If it were found out even if you decided not to prosecute for the few dollars you lost the employee and possibly the bank would be in position to be held responsible at the federal level. There’s been no crime identified for a scenario of a search engine being adjusted for political party bias.
I didn’t say what the state should do, although an unfair business practice charge would be appropriate. I just said I wouldn’t trust them nor do I see how anyone could.
Everyone promotes their politics
Switch from Google to DuckDuckGo.
Produces “different” search results … more conservative and less left-wing.
Check it out.
Any thoughts on Microsoft Bing?
I still find it vastly inferior to Google. Maybe it is the fact that a lot of my searching is technology-related, but Bing is constantly returning outdated results even for Microsoft technologies like .NET and SQL Server. Google does a far better job of keeping up-to-date. DuckDuckGo had the same problem when I used it.
Likewise; most of my searches are in the same domain and are apolitical. .Net, UWP, GCLIB, so on. My non-tech searches tend to also be non-political. Speaking simply from personal utility Google tends to work best for me.
Yeah, politics doesn’t factor in much to my searches either. Even when I am, it isn’t exactly hard to get search results to cover a wide range of opinions since you just have to look for the usual conservative and liberal sources.
That said, if I want to keep up-to-date on political news, I just use a news aggregator. Unfortunately, a lot of those seem to have started tailoring news to one’s bias. MSN was absolutely horrible about it, and Apple News eventually got as bad. Google News, ironically in the context of this thread, hasn’t done that, and their “Full Coverage” feature is good at offering various conservative and liberal sources.
This topic was automatically closed 14 days after the last reply. New replies are no longer allowed.