42 comments

[ 4.0 ms ] story [ 90.3 ms ] thread
The most interesting thing from my POV was the existence of http://x.ai/
Amy is a great assistant. But she has a long waiting list.
It's a shame they didn't minimally push the boat out for sexual equality by offering "adam" as a male alternative. Amy is (arguably) a stereotypically girly name and given that what's being offloaded here is administrative work, I think it looks bad to make it a female responsibility.
The posted article suggests that they already do allow this: Unlike some other intelligent assistant companies, X.ai gives you the option to choose a male name for your assistant instead: Mine is Andrew Ingram.
As far as the fiction can envision the techno future and human AI interaction, one of the more interesting books for me was SNUFF by Victor Pelevin. In it the main character is a freelancer drone camera operator living with a humanoid as a partner, tweaking her settings and getting very emotionally and sexually involved with her. And the devise is so human like and autonomous, that it eventually runs away from the owner leaving him with huge debt still unpaid.
Y'know, I have a terribly powerful desktop always running at my home; I'd love to be able to load a container on it which does all this processing locally.
Is there a (comparative) overview of all the tasks smart agents can do?
Enough with the 'woe humanity the machines are coming' pieces. After ten years of facebook, and three years of snowden leaks, I have a hard time finding sympathy for people who purchase 'smart' (read data gathering) devices from large conglomerates before considering the privacy implications. The cards are on the table; we all know what the terms of this kind of convenience are. If you're concerned for your privacy, then stop opting in for fucks sake. You've just voted with your dollars for a future of ubiquitous, autonomous surveillance. Again. Knowingly. The ignorance card is utter bullshit at this point.

The 'inevitable' future bewailed by these types of articles is only inevitable because this kind of 'shiny! buynow thinklater' mentality. To paraphrase Sartre, "we have the surveillance state we deserve."

If you're comfortable with the tradeoffs, I have no beef. I'm not saying that people shouldn't buy these things if they want them. And they are cool, wantable things. But it's 2016. You can have your thing or your soapbox. Both are respectable choices. But you can only have one.

I agree with you on the people who buy these things, but I think a more interesting question is how they affect the people who do not buy them.

Facebook is similar. I don't care one whit about Facebook users who complain that the company is using their information and that they have no privacy. That's what you signed up for, duh. More interesting is that Facebook amasses a lot of information even about people who do not use the service.

Right now the Echo is only listening inside of homes, so logically you can figure there's not much risk of third-party harm here. But this gets more tricky if these things start listening in public spaces, or if they're on phones...suppose the pizza delivery guy gets a cell phone that listens all the time, and it hears something interesting while the pizza man is at your door.

"Right now the Echo is only listening inside of homes, so logically you can figure there's not much risk of third-party harm here."

If I visit a friend who has an Echo in their home, I'm having my conversation listened to without my consent.

I wonder if we're going to start seeing a new norm (similar to "I have you on speaker, Jane is here too") around devices with always-on voice recognition?

You head over to your friends house and they hit the mute button on their way to the door, or ask Alexa to say hi.

If you're in someone else's home you have to assume he has listening devices. He controls his house.
You seem to believe that everybody somehow knows the consequences of "opting in". Does everyone have a CS degree in your universe?

> we all know what the terms of this kind of convenience are

Most people have no idea whatsoever what those terms are. Even among technical crowds I still find people assuming that humans are required for various tasks that have been automated for a long time. I seriously don't understand why you think people understand the "terms" of what big data and machine learning are doing to their data. Even simple things like the fact that cell phones give away your location can be a new concept for people that have only considered them telephones.

And why should most people have a realistic understanding of this stuff? The computer industry has been over-promising and dressing up their products since the transistor was invented. We have decades of services that dissemble as their business model, convincing people that their data is "private".

> 'smart' (read data gathering) devices

Why should anybody think that their TV is surveilling them. It wasn't long ago that saying your TV was spying on you could lead to a schizophrenia diagnosis. Nobody reads the legalese and manual for a TV before they bought it, understood it, and choose to trade their data. They bought a TV that advertised voice activation or some other feature. There is no reason for most people to think surveillance would be involved.

> stop opting in

While some people have started to realized how this stuff really works, the common response is to feel trapped without options. It will take time - decades - to properly educate the general public.

> The ignorance card is utter bullshit at this point.

Look at how many people here on HN that still think "anonymized" data cannot be correlated back to real their real identity. If people that understand terms like "hashing" and "INNER JOIN" are still figuring this out, the general public doesn't have a chance.

Hell, I'm the biggest cynic in most rooms, I border on diabolically clever when the mood grabs me, and I don't have the slightest clue what the actual consequences of opting in are.

But if the deal is too good to be true, it's only because you don't have all the facts. Information asymmetry is the very basis for capitalism.

"When I Google “kinkajou,” I get a list of websites, ranked according to an algorithm that takes into account all sorts of factors that correlate with relevance and authority. I choose the information source I prefer, then visit its website directly—an experience that could help to further shade or inform my impression of its trustworthiness."

While the Google experience is certainly more direct than the Alexa experience, you're still allowing Google to filter and order your search results. Check out the "filter bubble" if you haven't heard of it before.

The HN title, "Terrifyingly Convenient", hits the problem on the head. As long as technology users broadly continue to sacrifice security, privacy, and choice for basic convenience, these sorts of issues will continue occurring. There's a potentially very dark future ahead of us where all of our choices are made by entrenched mega-corporations who see us as nothing but a source of revenue.

"CONSUME!"

> There's a potentially very dark future ahead of us where all of our choices are made by entrenched mega-corporations who see us as nothing but a source of revenue.

How about corporations who see us as nothing but a source of data, not even as customers? http://www.faz.net/aktuell/feuilleton/debatten/the-digital-d... & https://vimeo.com/110222526

"While advertisers have been the dominant buyers in the early history of this new kind of marketplace, there is no substantive reason why such markets should be limited to this group. The already visible trend is that any actor with an interest in monetizing probabilistic information about our behavior and/or influencing future behavior can pay to play in a marketplace where the behavioral fortunes of individuals, groups, bodies, and things are told and sold. This is how in our own lifetimes we observe capitalism shifting under our gaze: once profits from products and services, then profits from speculation, and now profits from surveillance."

We're largely there already; it's just not obvious because it's incredibly ingrained in our culture. We've been there. One aspect of brands selling people an 'identity' (think apple or more profoundly, rolls royce). That's the consumer culture everyone hears so much about.

Companies have larger weight in American government than people do. Lobbyists and campaign contributions go a long way. Aaron Schwartz became dedicated to campaign finance reform for precisely this reason.

Part of the problem (and part of the solution to that problem) is that large scale companies must always optimize for money. They can care about certain things more, but still must optimize for money. Sometimes this benefits the general public, but sometimes it works against them too.

Decoupling industry from politics may be a good idea; we can give more weight to what is good for people than what is good for companies. Although politicians like Trump make me think that might not be ideal either :P.

I think the premise of the writer's concern is that people who would otherwise be doing more critical reading (looking seriously at the source of the information they're getting, exploring different sources, etc.) are just going to accept what's spoonfed to them. I don't think that's valid, because I think people are using Alexa only for very casual research - the equivalent of reading the box at the top of a Google query or the first paragraph of a Wikipedia article, or pulling a reference book off a shelf and reading a few sentences, in other words, cases where you wouldn't be doing much critical reading anyway. Voice interfaces don't make a lot of sense for doing serious research, it takes a long time to read text out loud, you can't skim, or skip things, etc. They're great for things like getting a definition of a dictionary word without looking up from your book, or settling a bet about the population of Ireland.
> As long as technology users broadly continue to sacrifice security, privacy, and choice for basic convenience, these sorts of issues will continue occurring.

(Note: I'm not attacking or critiquing your comment, just attempting to show another scope) I'll point out that the same argument applies for living in a city vs a suburb vs out in the country. It's a very very old dynamic, and sometimes people get ensnared, and we have to be careful. For example, countries where you need an exit visa to leave.

But like the meatspace equivalent, there's likely to be different situations simultaneously for different populations. With an ebb and flow between them and between more or less free choice.

Ah, the "filter bubble". It became terrifyingly real to me when I read a blog post by Dave Winer exulting in the fact that his "Fargo" program had reached the first page of Google results. When I googled it, his program was nowhere to be found on the first few pages. Hmm… I now use DuckDuckGo for most of my searching, and !g for almost all of the rest.
I always had imagined this convinient future to be about an intelligence that lived within my home and not a remote server farm. That way everyone had their own personal AI that grew with you and not connected to any others. There was no big brother since it wasnt centralized. Why cant a startup be formed to make this real?
Good AI requires huge amounts of data, like you would find on a server farm.
a single person produces enough data in my mind to train an AI for him or her. We just need to apply existing algorithms to the problem. We just need a lot of data to make a generic AI that works for all. Which isnt the case for the above proposal
The problem isn't the algorithms, the problem is the labeled training sets which those algorithms need to create models.

Also, to recommend news stories a la Google Now you need to scrape an appreciable portion of the Internet and extract features to score against your model.

So what you are suggesting this trend will be perfect as long as we control like keeping private home network? What a novel idea!!! However, in order for startup to be interested, this idea gotta have some commercial potential aka profitable. How do make this idea profitable though?
Not all innovation comes from startups. There are probably lots of people who would be interested in working on a free, open source AI system if someone were to organize such a project. It would be interesting to see who the "Linus Torvalds of AI" will turn out to be.
So you can get voice recognition at home, but you end up losing updates, new features, and ability to interoperate with other services. Cloud based services make it so that one person doesn't have to carry this load themselves.
Can't I get those things from `sudo apt-get upgrade`?
Having mass market B2C customers do this is a customer support nightmare.
Standardization makes it so that nobody has to carry this load themselves.
In your car, OnStar listens to you. That was the first widely deployed "always listening" system of that type. On January 1, 2014, GM changed their privacy policy to allow them to use any info about what your vehicle is doing, including where it is, for marketing purposes.[1]

At home, Echo listens to you. Amazon's is vague about how much they listen to.[2]

And, of course, there's the XBox 360. It sees you when you're sleeping. It knows when you're awake. It knows if you've been bad or good.

You'll never be alone again.

[1] https://www2.onstar.com/web/portal/privacy

[2] http://www.amazon.com/gp/help/customer/forums/ref=cs_hc_g_tv...

> You'll never be alone again

Or, if this bothers you, you can choose to avoid the hedonistic treadmill altogether and not outfit your house with an Alexa, a Kinect, and countless other mostly useless devices that people lived happily without 10 years ago.

The problem with this nice theory is you'll be getting it whether you want it or not. As revenue comes in from the sureillance society, manufacturers will (and alrewady are) crammin the functionality into devices. You, as a customer will have no choice: sooner or later you'll need a new fridge, and a spy fridge will be your only option.
That is simply not true. Maybe the latest and hipsteriest devices will all have those features, but the market is vast and caters to all sorts of humans and price points. You just have to care enough to make that choice as a consumer.
Try to buy a high-end TV without Wifi/Smart/3D useless anti-features.

It's getting harder each day.

And you'll have no choice but to connect it to your network? Or no way to interfere with whatever wireless tech they use ? Seems a little far fetched
There's a few smart TVs that will complain if they haven't been connected recently, so I don't think it's that far-fetched.

That's before considering that the newest version of the BluRay spec mandates network connetivity to authenticate the disk you're playing. Which is a more special case of stupid.

Soldering iron solves a lot of "convenient features"
I think this problem is on the verge of solving itself again. It'll come back by the end of the 2020's of course, but we are at least due for a respite.

There's so much work being done on scalable server tech right now, and enough broadband to use it. O little nudge and we'd be close to having a server appliance that people could have in their own homes, and call from their mobile devices.