[TomFrost]

The problem I've always had with audio interfaces is that input is not private. Requests on public transportation are heard by many. Requests at home turn into a conversation with the roommate, spouse, or children. Requests walking down the street make others question the mental wellbeing of the person talking to him/herself.

I'm reminded of the Ender's Game sequels in which the protagonist wears a small earpiece with an AI named Jane. He communicates with Jane by "subvocalizing" -- mentally saying the words, physically barely uttering a sound. The AI understands.

A few years ago there was a TED talk (forgive me; unable to find the link) on which a technology was demoed to do something similar. Sensors placed around the throat, combined with EEG sensors around the temple, allowed a man to transmit text to a computer by following all the mental and muscle processes of speaking, stopping short of moving his lips in an obvious fashion or making sounds. The sensors allowed the computer to translate their input to actual words.

Perfecting and miniaturizing that technology, then combining it with an in-ear AI, would be a game changer.

[melling]

Yeah, you can always find a use case where something won't work then spend your time discussing why a solution won't work for you.

However, I imagine that a lot of people would love the solution in many scenarios. Personally, I have no problem asking my phone a question when I'm walking down the street or driving. Basically, it would be nice to have the imperfect solution now then take it from there.

[lewisl9029]

Agreed regarding the issues with audio interfaces.

I'd personally love to experiment with a device with a Google Glass-type display, eye tracking to function as a virtual cursor, and a compact ring-like device to enable interactions with that cursor (such as tapping, dragging gestures, etc).

[rmac]

The future is brain-to-text, where our thoughts get piped directly to Alexa/Siri/Cortana/Now

http://www.kurzweilai.net/brain-to-text-system-converts-spee...

No need to utter words in public.

[erikpukinskis]

Maybe. I support brain-computer interface research, but I don't think you should be so certain that it's the future.

Imagine you could stimulate the brain directly, what would that look like? If it's visual input, it's going to look a lot like stimulating the visual cortex, right? So you're suggesting we will invent some sort of biomedical device that can project an image onto some region of cortex. That would be neat!

But isn't there already a device that can project an image onto a region of cortext? There is! It's called the eye! What is really gained by stimuating the visual cortext directly, vs just having, say, an eyebrow piercing that just does the same thing by coming in through the eyeball?

Well, ok, maybe you'd like to get a little more bandwidth than the eye can provide? But I'm not sure that really makes sense. The eye takes in a LOT of bandwidth, and most of the time, unless we are fighting a bear or trying to run a football through a bunch of bodies, we're not even using all of the bandwidth.

Maybe you want to be able to look at the world and get input subconsciously without messing up your beautiful view. But you can just dedicate a some small number of "pixels" of your visual field to non-visual data. If your brain can't integrate that data with the visual field, it will just become a blind spot, and it will be as invisible to you as your current blind spots are.

Really, our perceptual system could be much more powerful, but beyond a certain point there's no reason to have more bandwidth because the existing stuff is all a brain of our size can really synthesize in realtime. You could add additional inputs to your brain, but your brain will just grow around them, dulling your connection with other parts of your body. That's not necessarily bad, but, again, why not just use those parts of your body for the input?

The planet Earth has spent about a billion years trying to design the perfect input/output interface for the human brain and it's done a pretty good job. I'm not saying we can't improve on it, I'm just saying it's going to be hard to do any optimizations at the hardware level. And we don't really need new hardware to plug in to computers, because our brain is already happy to remap itself to virtual inputs! Pretty much all of the things people imagine doing with a BCI can be done more efficiently, and less obtrusively with augmented reality.

[Ludo0]

As a deaf dude who has spent his whole career working on software I see it as a software problem.

Despite all the advances, we still don't know how to encode sign language.

[Odin78]

If you send the input through the eye it would still appear in your regular field of view, right?

But what if you want to keep looking at whatever you're currently looking at without obstruction?

If you project the input directly onto the brain you could have two separate images, similar to how you can create a picture of, say, a bear in your head while still looking at your computer screen.

[chongli]

Imagine you could stimulate the brain directly, what would that look like?

It needn't 'look' like any of our senses. Direct brain-computer interaction could take the form of an internal dialogue, similar to how a person might mentally rehearse both sides of an anticipated conversation in their head.

[erikpukinskis]

OK! So you're talking about sound data then. We also have a really great input device for that.

How would the direct neural input be better than hearing?

[chongli]

It is not sound data. It is words without sound. There is no tone, no pitch, no volume. It is just words. Thoughts. Concepts. Ideas.

Hearing has so many drawbacks. It is limited in range. It is affected by environmental noise. It is linear and if you miss a word you need to have it repeated somehow, which may not always be possible. Thoughts have none of these limitations.

[XorNot]

Infinite volume with no hearing damage?

I've never been able to find a satisfactory answer on how the cochlear implant behaves like that. I somewhat suspect nobody with hearing who developed it actually knows.

[erikpukinskis]

You mean like this? https://en.wikipedia.org/wiki/Up_to_eleven

[ferongr]

If we have the interface why stop at brain to text? Why not make the computer a part of our brains?

[raddad]

>If we have the interface why stop at brain to text? Why not make the computer a part of our brains?

This is how the Borg came about.

Actually I read about some guy who has a chip implanted which allows him to use his otherwise unusable hand. Right now it requires a bank of computers to function in the lab.

Paralyzed man with breakthrough brain chip plays guitar video game, swipes credit card

http://wexnermedical.osu.edu/blog/new-tech-helps-paralyzed-m...

[dsr_]

Two major reasons: upgrades and security.

Do you want to get the shiniest, greatest interface of 2027 installed, only to discover in 2035 that after three upgrade cycles, your hardware is obsolete and won't be supported?

Do you want to get the best-tested, most upgradable interface of 2027 only to discover in 2028 that you were one of the hundred people who were hit by a zero-day exploit and a bot-net has root in your head?

[userbinator]

where our thoughts get piped directly to Alexa/Siri/Cortana/Now

...and around the same time, they'll probably also figure out how to pipe adverts directly into our brains. I guess you could call that the scary side of "seamless" HCI.

[xirdstl]

Maybe, but that has to be a long way out. I can see what the parent describes happening well before brain-to-text.

[amasad]

I subvocalize when I read and it feels like I'm breathing the words. I think my vocal cords move too.

[ndesaulniers]

Jane! Just bought my little sister Speaker For The Dead. \m/

[notliketherest]

You are forgiven in the lack of (Ted Talk) reference, oh wise Hackernews commenter.