Forget meagre voice commands for iPhones, these guys claim to have come up with a way to control them with their minds. Did these two random dudes just accomplish something that's defeated scientists for decades? Or is it all bullshit?
Project Black Mirror is in the very early stages of development—these two guys, named Ollie and Josh, just came up with the idea over Halloween weekend, but holy crap they claim to already have a working prototype. Here's how it supposedly works, from their blog:
1. ECG pads provide raw skin conductivity / electrical activity as analogue data (0-5v).
2. This is plugged into the Arduino board via 4 analogue inputs (no activity = 0v, high activity = 5v).
3. The Arduino has a program burnt to it's EPROM chip that filters the signals.
4. Josh trained the program by thinking of the main Siri commands ("Call", "Set", "Diary" etc.) one at a time and the program where we captured the signature brain patterns they produce.
5. The program can detect the signature patterns that indicate a certain word is being thought of. The program will then wait for a natural ‘release' in brain waves and assume the chain of commands is now complete and action is required.
6. The series of commands are fed to a SpeakJet speech synthesiser chip
7. The audio output of which simply plugs into the iPhone's microphone jack.We still can't believe it really works.
Yeah, I kinda can't either. In fact, it's entirely possible (likely even) that this is bogus and I'll look like a tool for posting this . Call me a rube, but I want to believe this guys. I like their style, with blog entries like, "this shit might just work!" Here's a video of one of their first "successful" tests.
So, obviously, the system is big, bulky, and overly complex right now, requiring a separate computer and a bunch of hardware. How impractical it is right now is insignificant. Here's how it could work (this is just my theory): The ECG pads connect connect to a small, pocket-sized device. That device connects by Bluetooth to the iPhone 4S, and an app on the phone translates learns your patterns and translates the signals to Siri. Something like that. This could also potentially solve Siri's problem understanding people with accents.
These guys just got approval from Kickstarter to launch a project to fund this gear. Keep checking back for that. If this is real, I really want to see it happen. If it isn't real and they actually try to launch a Kickstarter Project on the strength of a phony video, it would be pretty major scam and they'll probably end up in jail. I'm hoping, for everybody's sake, that it's real. What do you think? [Project Black Mirror via Ubergizmo]
Update: Oooh geekfight! It's looking more and more like our skepticism is well-warranted. Trevor Coleman of InteraXon (a well-respected company in the field) had this to say:
I'm one of the founders, and the COO of InteraXon (www.interaxon.ca) a thought-controlled computing company in Toronto, Canada, and world-leaders in experience design for brain-computer interfaces.
I think this looks, to be blunt, impossible. And I have doubts about its authenticity.
For one, the way you've described EEG data collection doesn't seem relate in any meaningful way to how brainwaves are actually collected. If you've simplified for the lay audience, perhaps you'd care to elabaorate to substantiate your claim.
Also I'd be very interested to see a demo where the system does arbitrary commands on video at the prompting of a neutral observer.
If I'm mistaken, and you've made some extraordinary leap in the technology then you have my sincerest apologies.
If that's the case, we'd love to know more about your work and will be more than happy to encourage you in success as you join us and other companies that are working to build a thought-controlled future.
If this is a hoax, you should know that you are damaging the work of a great many people who are doing their best to bring this revolutionary technology to market. I'd certainly advise you to make sure you can back up your claims *before* you start taking people's money.
Regards,
-Trevor Coleman
So there's that. And—and this is harder to quantify—but there's something about their delivery in the videos. It kinda seems like they're acting, no? And not all that well. I'm afraid I've gone from skeptic to "bullshit caller". If I'm wrong, then I apologize to these gents, but we're gonna need some much better proof.
UPDATE 2: Further damnation. This time from Greg Courville, an undergraduate physics student at the University of California:
If you haven't spotted it already: not only does there appear to be virtually nothing connected to the SpeakJet chip, it's also placed sideways in the breadboard, which shorts a number of critical pins together, and therefore would make it impossible to use.
The project's blog claims that "ECG pads provide raw skin conductivity / electrical activity as analogue data (0-5v)". Let's go ahead and ignore the fact that they seem to have confused EEG with ECG (which is a completely different type of measurement), and the spurious reference to "skin conductivity" (ditto). The most glaring problem with that statement is the claim that EEG signals fall in an Arduino-friendly 0-5V positive voltage range. Suffice it to say they don't.
What did Project Black Mirror do? Based on their written statements, their YouTube videos and the photo shown below, they connected the electrode leads directly to the analog input terminals on the Arduino.
In short, there's not even the slightest chance that they are actually recording EEG signals this way.
You can keep up with Brent Rose, the author of this post, on Facebook, Google+ or Twitter.