I like to think that Google is just not a conventional company and that they do everything to not become one. They creatively solve challenges (like how to keep certain cookies when a user presses "remove all") and this helps them provide unbiased, accurate, and free access to the world's information. People wanting to stay signed in can rely on them to do the right thing, even if in the near term the financial returns are not obvious. Their corporate structure is aligned with this objectivity: Serving their end users is at the heart of what they do. When people are logged off you can't really consider them your number one priority, because they don't either. Many Silicon Valley companies are under pressure to keep their earnings in line with short-term analyst's forecasts, Google is no exception here. But forced and universal log-in to Google's services optimizes for long-term return. Everybody hoped, but nobody believed, that Google would end up in this powerful position: Being able to strong-arm your power users, while still providing the best, most accurate, search results for people not logged in. In the end, this business tactic of universal and forced log-in makes the world, and by extension the internet, a better place. I know some people here raise privacy concerns, but this change clearly protects a user's privacy.
If the police can tie the IP to a person, then it "can be tied". Some courts decided before GDPR that IP addresses can't convincingly be enough to tie a person to a crime committed with that IP address, but that is not to say it isn't PII, or can't be tied to a person.
It depends on the purpose / if you have a legitimate need to store them.
Are you storing IP's for a certain time, so you can discover and investigate attacks?
> include preventing unauthorised access to electronic communications networks and malicious code distribution and stopping ‘denial of service’ attacks and damage to computer and electronic communication systems.
Are you storing IP's for a certain time, so you can run analytics?
> Further processing for [...] statistical purposes should be considered to be compatible lawful processing operations.
Are you doing direct marketing to your customers?
> The processing of personal data for direct marketing purposes may be regarded as carried out for a legitimate interest. [Legitimate interest] There is a relevant and appropriate relationship between the data subject and the controller [...], such as where the data subject is a client or in the service of the controller.
If you are not big enough to have a law department look at this, you don't need to worry so much about a weblog.
Stay on GMail, but don't log-in to the web interface. Instead use Thunderbird. That way you can read your mail without being forced logged-in to search or video.
Google has been making this harder and harder though (now calling Thunderbird insecure, and requiring you to allow insecure app's in the options), but it is a good canary in the coal mine: Once I can't avoid this unified/universal log-in, I'll know Google turned me into an adversary.
With this setup I deem Gmail to be the best service for privacy and security.
It was never removed from the code. You seem to think it was, and that the media attention made them add it back.
Try to find a snapshot of the code that does not include it (I can only find snapshots with it, so that won't suffice as proof).
About Eric Schmidt (who I won't defend) the full quote and context ("a casual jokey interview"):
> "The idea was that we don't quite know what evil is, but if we have a rule that says don't be evil, then employees can say, I think that's evil," Schmidt said. "Now, when I showed up, I thought this was the stupidest rule ever, because there's no book about evil except maybe, you know, the Bible or something." In the end, though, he believes it has worked, by giving employees a way to point out things they find unethical.
Subtle but it is there: The former CEO does not think that doing no evil is stupid, he thinks using a rule like that, without properly defining evil was stupid. Then he changed his mind.
Compare with the JSLint license that states: "the Software shall be used for Good, not Evil." You can find that rule stupid and vague, while acting like Mother Theresa.
> "As Google (and some others) interpret it, this additional requirement constitutes a vague use restriction and thus makes the license non-free. Chris [DiBona] explained that if I were to remove that line from the license and 'return to a proper open source license that we support,' then jsmin-php could stay on Google Code.
The disagreements were about strategy and mobile experience, so not enough to call someone poisonous. Place Yahoo, AVG, and Moz Corp against Brendan's co-founder status and JavaScript.
The other toxic drama was about a 1000$ political donation. Very sad to see an unpopular view used to paint someone as a monster. If democratic political views are enough to kill someones career, it is more a sign of the toxicity of today's social media and activism than a character judgment.
> Does one really want to associate themselves with Brendan Eich though?
No, not really. Not a good political idea to align yourself with the black sheep. With all the tars and feathers he just looks weird.
Though it also works in his favor: Brave is popular amongst the alt-right and technical-minded early adopters. They see it more as character assassination and SJW corporate culture taking desperate vengeance for Trump's win.
The data is correct and a-biased. If you ask 100 people around you, they are, on average, more likely to have had a negative burrito experience, than a negative pasta experience.
The learning algorithms are crude and dumb. They will simply fit to any data you provide it (you choose how many Mexican restaurant food reviews you train your sentiment classifier on). Then they count how many times the words "mexican" and "man" and "mexican man" appear with a positive or negative label in the train set. And objectively try to give the best probability for that.
Current sentiment analyzers are not AI: no common sense, no understanding, no reasoning. We are just rushing to replace looking a job candidate in the eyes with running some 1960's logistic regression over their cover letter. Let's hope for their sake they did not manage a Mexican restaurant.
> A computing professional has an additional obligation to report any signs of system risks that might result in harm. If leaders do not act to curtail or mitigate such risks, it may be necessary to "blow the whistle" to reduce potential harm.
Leadership was actively advancing the project and promoting obscurity / secrecy. Sundar did not act to mitigate risks of harm to free information and dissidents. People were not given enough clarity to make good moral decisions.
The entire project reeks of a top-down ethics violation. You can't with a straight face introduce AI ethics guidelines, while you have backdoor meetings with need-to-know engineers building a surveillance and information manipulation system.
An objective party within Google should work hard to protect Google's values. To me, an outsider, Sundar can't be trusted anymore on responsible ethical AI (and by extension: AI itself). Probably some misaligned incentives there.
> As a leader in AI, we feel a deep responsibility to get this right.
So get it right. Start by fixing the wrongs and keeping consistency with your messaging.
Or tell me how the planning of a opaquely censored, dragnetted, privacy-intruding, and authoritarian-friendly search platform is consistent with:
1. Be socially beneficial.
2. Avoid creating or reinforcing unfair bias.
3. Be built and tested for safety.
4. Be accountable to people.
5. Incorporate privacy design principles.
7. Be made available for uses that accord with these principles.
We will work to limit potentially harmful or abusive applications.
We will not design or deploy AI in the following application areas:
Technologies that cause or are likely to cause overall harm. Where there is a material risk of harm, we will proceed only where we believe that the benefits substantially outweigh the risks, and will incorporate appropriate safety constraints.
Weapons or other technologies whose principal purpose or implementation is to cause or directly facilitate injury to people.
Technologies that gather or use information for surveillance violating internationally accepted norms.
Technologies whose purpose contravenes widely accepted principles of international law and human rights.
The only thing consistent with the AI ethics guidelines (a plan going forward, already abandoned on release) is the pledge to technical excellence. I am sure, as the leader in Search, that Google is able to build a fine custom solution for the Chinese government.
> 4.1 Uphold, promote, and respect the principles of the Code.
The future of computing depends on both technical and ethical excellence. Computing professionals should adhere to the principles of the Code and contribute to improving them. Computing professionals who recognize breaches of the Code should take actions to resolve the ethical issues they recognize, including, when reasonable, expressing their concern to the person or persons thought to be violating the Code.
> 4.2 Treat violations of the Code as inconsistent with membership in the ACM.
Each ACM member should encourage and support adherence by all computing professionals regardless of ACM membership. ACM members who recognize a breach of the Code should consider reporting the violation to the ACM, which may result in remedial action as specified in the ACM's Code of Ethics and Professional Conduct Enforcement Policy.
Please voice your concern. Your voice as a Googler is louder than mine.
> thefacebook.com allows for targeted advertisement on the basis of any (or a combination of) the following parameters:
> College/University, Sexual Orientation, Degree Type, Zip Code, Concentration, Dating interests, Courses taken, Personal Interests, Class Year, Clubs and Jobs, House/Dormitory, Political Bent, Age, Number of friends, Gender, Site Usage.
So blatant targeted advertisement has always been part of Facebook. Lots of (now wealthy) investors and early adopters had no problem that its users were being targeted on the basis of their sexuality.
The market will go where it wants to go: Instead of explicitly targeting men, you now target people that liked WWE, Family Guy, study business school, and searched for "hackernews". Or you make the advertisement itself appealing to men (and take the small loss on an irrelevant audience).
To me it seems all the work of the same spammer(s). In such a case, do some manual intelligence and wrap it up. It won't scale to all forms of spam, but if a simple regex can uncover 250k+ results in 10 minutes, a manual spam fighter can still block millions of pages a day (and warn the webhost, remove these flakey ads from their networks, etc.).
No doubt the recent machine learning hype has given spammers more advanced tools to avoid detection.
Actually, shortly after I learned that Google switched to neural networks for search results, I noticed an old style of spam making a huge comeback: Either markov chain or neural network generated text, taking from related websites (I suspect they look at the top 10 websites for certain valuable terms).
This gibberish actually outranks legit content which refers to my content, sometimes even my own articles, especially when it is turned into a PDF.
Seems like it is easy to block ~250k webpages like:
- Two main clients did not pay bills on time, causing me to run out of runway. Getting them to pay became the single most important issue, and after they finally did, I was stuck with a 2.5 month backlog, two important clients who couldn't be trusted to pay on time, and a broken spirit.
- Created a b2b app, spend too much time on making it work great, not nearly enough time/effort on marketing it. Found that the people who could understand the technical superiority were not our target market, but just people who estimate technical superiority from marketing. Couldn't and wouldn't compete on that. Then we landed a big client who paid us handsomely to lose our focus and turned us into a non-scalable consultancy business which we sold for an almost symbolic value.
- Used networking to get together a team of 5 researchers/engineers in a hot field. Found funding prospects, but only on the condition that they pay us (slightly below market average), but the equity would be in the single digits (even for me). So this blew up, and a few months later the most promising researchers got jobs at FAMG. I think most $$ potential was lost here (would have been an easy acqui-hire).
Learned:
- The cheaper the price, the cheaper the customers. I took pride in offering great customer service, but spend hours looking into issues for 15$ accounts (would have been cheaper to fire them).
- The smaller the startup, the lower you are prioritized come pay-day. Would be more aggressive in collections/SLA.
- It is not how good your product is, it is how much people are willing to spend for it (how good they perceive it to be). Non-technical people are easier convinced by a sales call, than an impressive demo.
- Learned how to deal with financial stress and to keep a good sleep/workout schedule, else my output drops way too low.
- 15% equity of 0$ is 0$, but you also can't expect the really good people to work for your startup without any potential upside or throwing them a bone. Strike while the iron is hot.
I think the fundamental difference is null. Or maybe I am just one of the first to start to say: Hey! That's wrong!