Live data from Hacker News

Computer, Respond to This Email

googleresearch.blogspot.com

41–50 of 202 posts

Re: Computer, Respond to This Email

#41
Minus all other concerns, for the operation / infrastructure / devops, the auto-replying about server downtime or inquiry about performance, those kinds of stuff is something I am actively thinking about and starting to learn how to implement in my organization.

Re: Computer, Respond to This Email

#42
Sounds like the system has no way to take into account context. Each message is encoded into a 'thought' vector, but is it difficult to come up with a 'thought' that changes meaning depending on prior messages? I imagine it is such a concern that leads one to look into Hierarchical Temporal Memory and similar techniques.

Re: Computer, Respond to This Email

#43
post #26
post #24

Earlier quoted context omitted.

Stop with this "you're the product" nonsense. Please.

Stop making the user the product. Please.

I have somehow managed to avoid being sold by Facebook.

They have managed to sell the ability to include content in HTTP responses sent to my requests, though.

Re: Computer, Respond to This Email

#44
post #6
post #3

Finally! This will dispel all those rumours that Google reads all your email.

Why do you care if a computer reads through your email? Its not like Larry Page is looking through your search history. AI seems to just be data expressed in some way, ei me and you are a physical expression of years of gigabit streams of data. This will only become more strongly felt as we get closer to serious AI. Sadly I think its something you either get used to or go live in the woods.

Same argument as the NSA: "It's not surveillance until a human looks at the data"

Re: Computer, Respond to This Email

#45
This looks more powerful than iOS's context sensitive keyboard, but that also looked impressive when demoed. It will be interesting to try this out with real-world emails and see how well it works in practice.

This gets really exciting when implemented on a smartwatch. Single tap responses that are actually useful would make smartwatches significantly more powerful. Speech recognition is great but it's not always appropriate for social environments.

Re: Computer, Respond to This Email

#46
post #26

Earlier quoted context omitted.

Stop making the user the product. Please.

He's not wrong and you are. You are not "the product", and saying such is so divorced from the truth to be nothing more than a lie at this point. "You" are not being sold, access to you is. Look at some ads, get some great web services. Seems a fair tradeoff.

I'm not entirely sure this is a semantic argument worth wading into, but I think there's a wrinkle worth addressing in the "you're the product" mentality. In order for Google to effectively serve you ads (and convince advertisers it's worth their prices), they accumulate vast amounts of information about their users. From some perspective one's identity is largely based on or equivalent to one's purchases, tastes, interests, friendships, knowledge, etc. Advertisers (like Google) try to know as many of these as possible about everyone they can, and I can see how for many people this is equivalent to Google selling "you" -- they have a digital representation of your identity, and their ability to comprehend it is the real service that sits behind the advertising.

Re: Computer, Respond to This Email

#48

Earlier quoted context omitted.

He's not wrong and you are. You are not "the product", and saying such is so divorced from the truth to be nothing more than a lie at this point. "You" are not being sold, access to you is. Look at some ads, get some great web services. Seems a fair tradeoff.

I'm not entirely sure this is a semantic argument worth wading into, but I think there's a wrinkle worth addressing in the "you're the product" mentality. In order for Google to effectively serve you ads (and convince advertisers it's worth their prices), they accumulate vast amounts of information about their users. From some perspective one's identity is largely based on or equivalent to one's purchases, tastes, in…

I disagree that this is an argument over semantics seeing as how "Google is selling access to your attention based on what they know about you" is much less scary sounding (and more accurate) than "Google is selling you/your information".

The latter makes it sound (willfully) as if random third parties have access to the information Google has collected, which is a pretty blatant falsehood. Anyone is able to go sign up for an Adwords account and see how "targeting" works. At no point does the advertiser get to see anything about the people.

The whole "Google is selling you/You're the product" meme is a breathless, thought terminating cliche that needs to die.

Re: Computer, Respond to This Email

#49
post #37
post #34

Earlier quoted context omitted.

> * Stop with this "you're the product" nonsense. Please. * what would you call it?

The product is the service being offered, in the case of Google it's search, mail, maps, etc. How they make money is unrelated to what the "product" is. E.g. I switched from free ad-supported YouTube to the paid ad-less YouTube Red, does that mean what the "product" is suddenly changed? Also you really are the "customer" or "user" of the many services being offered by Google just like an advertiser is a "customer" of…

This is a lot of words that amount to "Nuh-uh!"

The entire point of the "if you're not paying, you're the product" spiel is that everybody already knows everything you said, but there's a fundamental difference between paying and not paying. You're not going to get that far demanding that people stop thinking there's a difference, nor are people going to stop verbalizing that difference, nor should they. What's cliche on HN is something the vast majority of the rest of the world still has not heard, and in this case, ought to.

Re: Computer, Respond to This Email

#50

While the technology is very cool, this pushes me that much closer to a "dumber" email service. It's always a conflict for me as I love shiny new things, but I'd rather we let AI loose on someone else's email, especially when the AI's revenue stream is advertising (a business predicated on knowing as much as possible about your audience).

I don't think I understand your reasoning. You're already tolerating an AI processing your email to produce the ads, so assuming that, why does a cool new feature push you closer towards a "dumber" service?

how do you know the poster isn't blocking the ads? (or for that matter using a separate client?) and anyway, ads are stupid awful crap, so who cares if they get that part right? It's nothing like actual email messages.
Post reply on HN