Given the “insights into your life” feature, maybe it will refuse to record “new” voices without an explicit check? The implication of the marketing copy is that it understands both the words and the sentiment so it wouldn’t be that absurd to think it can also distinguish one person from another. I assume in the end Rewind’s expectation boils down to “trust us” which is not good enough for me but probably good enough for them to make money for people who don’t care.
I want one that has a print of the "On Air" signs seen in studios.
I honestly think that something like that needs to happen to things like the Meta RayBan glasses that do live streaming. Instead of the eyeballs that Apple's device is using, it should have a "Recording In Progress" or "On Air" across the lenses. Just because I'm in public and prone to be captured in someone's image capture doesn't mean that I have to be part of their money making venture.
i've been looking at the idea of ball cap with surface mount LEDs under the bill of the cap, but i'm not much of a cap wearer any more. i then thought if they can fit cameras into sunglasses, why not do the Orbital thing with bright LEDs attached to my sunnies? then, the evil grin comes out and starts thinking more sinister with lasers to not just ruin the shot, but the sensor itself. then i realize, i've got work to do, and forget the whole thing
If you mean the retroreflective stuff, that's because it requires a bright flash to work. Modern digital cameras now are generally good enough to not require flash for a good picture, even in low-light conditions, so paparazzi don't use them near as often.
Clothing or accessories that shine IR are a completely different functionality and while they could be stopped by paparazzi buying better IR filters, that requires an additional cost on top of the camera
about 10 years ago I spoke with an officer who knew the law for our 'one party consent (aware) state' and was told then that you can legally record others' voices with zero person consent if it was 'posted or announced'
- However the way the law is written, it's vague enough that the 'posting or announcing' you need to give notice if recording others not in your conversation could be:
simply having 'we're recording' printed and posted on the door of a club would be legal, (nothing saying how large of print, or if it was surrounded by 100 other printed things)
and so would someone shouting 'hey we're recording' in a bar .
- that's the way the law was technically written, and though we know that yelling 'we're recording' would not be heard or even register to those who did hear the statement, that and what it would mean and most would miss the printed notice on door or forget it - it would technically be legal.
(no need to do both, either would satisfy law as written)
Not sure if the laws on this have been updated since then, and pretty sure it varies state by state.
I am not a lawyer, and am not referencing any case law here.
I had the same thought. As far as I can tell, they are leaning on the ability to pause recording at any time, meaning that in all-party consent states, it's up to the user to disclose that they are employing a recording device. That seemingly defeats the point of the device in many situations where the default state should be off. I would most benefit from something like this in public situations where I'm meeting a lot of people, but it'd be impractical to use it that way where I live.
That would be disappointing. The way I've always figured this would work is to just record the wearer using voice recognition and filter out everyone else unless some consent phrase is heard.
That would be the ideal way, but from the use cases on the site it doesn't seem to be the intended use.
My prediction is typical move fast break things tech bro culture. Release a product to everyone that is illegal in most states, wait for people to get hurt by it and care enough to complain, withdraw the product from the problem areas.
The fact they're so vague on who is recorded means they don't actually care, or at best don't have a solution yet to the problem.
>> We offer features for you to ensure no one is recorded without their consent.
> it is fundamental to the product experience
Without speaking too much to the underlying implementation, could you at least:
Clarify if "fundamental to the product experience" and "ensure no one is recorded without their consent" means the control loop is able to sense a non-consenting speaker and stop recording?
Clarify if "recorded without their consent" means never sending recordings off-device? Or does it just mean omitting the record from user facing long-term storage?
Clarify whether consent is opt-in or opt-out?
Clarify whether consent is actually consent, or the effective equivalent of supplying the user a "resume recording" button?
FWIW, I’m an existing customer (of the app) and this has significantly damaged my perception of Rewind as a company. You should stop talking about it until you have a credible answer to this question.
Exactly. Their desktop app uses whisper.cpp for handling transcriptions, then stores the output in a SQLite database. Summarization is handled by piping the transcript into GPT-4…
I wonder how it handles noisy environments, low volume when talking, and TV or music playing in the background. I've tried using pocket recorders while driving and the result can be unintelligible. Would be more useful if it could do a rough interpretation of activities (based on sound or accelerometer) to make it easier to place transcriptions in context. Something like <music playing: 45 minutes> <walking: 3 minutes> <door opening: 1 minute> <transcription of conversation: 10 minutes>
Reminds me of Deborah Estrin's work on Participatory Sensing:
I remember a conference meeting where her team wore cellphones around their necks, set to automatically capture images every few seconds and didn't reveal it until they presented a paper about it, including just-captured images of all of us! Everyone was a bit surprised, some people got upset and I think she promised to delete all the images. She's done some work on recording diet with this kind of periodic camera capture also. Audio detection of eating is possible too. A life diary would be something many people find useful, especially those with memory issues (which is just about everyone to some extent). The privacy implications are severe though.
Time to start building an AI pendant that is active attacks against surveillance that broadcast distortion and garbling above and below the human voice level to harmonically distort recordings in a known space without disrupting human communication. You can overdrive tiny speakers or disrupt known NLP algos, both, whatever.
Or maybe every house needs a Cone of Silence like in Get Smart.
A long time ago I had an idea after leaving DEFCON about making a standard gizmo, something like a lapel pin that connects via whatever to whatever supercomputer someone is carying in their pocket and that thing presents an obvious color code. Lets say 4 quadrants with a circle in the middle. I think ideally it would be like epaper with maybe a backlight option. Non intrusive but could be vivid. This could let people know in polite society if they did not want to be recorded, were recording, how they were recording, etc. Obviously this doesn't stop anyone from doing whatever, but you can monitor signals from people and if someone is doing something they say they aren't that can be shared. You can also transmit information like PGP public keys via ultrasonic handshake or whatever.
I mean, day to day I could see never needing it but going into the office, taking the train, going to court, going to a concert, there are a lot of ways something like that could be very useful and if adopted at any sort of scale could error check and blacklist bad actors.
I've been working on things that basically do that for a decade. I mean I didn't really think of this specific application initially but I did kind of.... rotating parabolic ultrasonic arrays something something. I'm looking for partners and investors!
I recently threw up a hackaday for my development cyberdeck and am sifting through all the stuff from over the years to finally get this to market. With the new CloudFlare AI workers, a whole bunch of the infra I used to have to maintain is a moot point so I'm looking to really hit it hard in Q1 2024.
Yeah. I'm working on that. This isn't really a pitch and you probably wouldn't want my cyberdeck. It's my dev workstation and yeah, no one wants that. I mostly mentioned it just because I'm getting to a phase where I am interested to get eyes on the weirder parts. Honestly, the concept itself is kind of a filter. If you look at it and get why you would want it, then we should probably talk, you're likely doing some weird stuff yourself!
The real pitch is working towards seamless natural user interface for streaming composite spatial data. The easiest thing to explain that people will get is architectural pre-visualization. You've got plans and lot lines, input plans, it generates a structure, that's a known solution, easy peasy. That is layered over GIS data from whatever source. Construction is underway, I can do drone surveying and flightpath automation to do photogrammatry or NeRF or whatever to build and overlay the model from the scan. Simulations are easy for erosion, or looking at the light at different times of year. I can drop ship you a 3D printed architectural model or some sort of widget. If you point your phone at that widget, you can interact directly with the information. Standard AR fare, you can go onsite and see the AR stuff. Builders can take scans of things in production and combine them. Yadda yadda. I have a company that is selling basically that. But to get to there, what I'm after is real time streaming composites with effortless and inexpensive mcguffens.
Composing things for a real time stream to be delivered for people to look at needs to have sort of LOD fallback. Streaming compressed point clouds that can be altered in real time, cameras that can be controlled or tabbed through, pulling video from multiple sources and muxing all of that into a cohesive digital product attached to a physical / inexpensive medium that a person can interact with. To build these things I've also had a focus on audio reactivity. Composition based on the sounds produced in the real world and in the digital world. Say 20 people have the same mcguffin, they can all fiddle with parts in place on their desktop as an AR or VR experience, or they can just open up a browser and mess with the world like a traditional video game. Or there can be a stream of 4k to Youtube or Twitch or whatever and that can be synced up to augment.
This is a long term study on what I don't like about modern computing and what I want to see. The stuff that makes money is high throughput distributed GPU compute and GIS mapping. These days people also want LLMs in everything so I've got the stack to produce that and it pays the bills.
The parts are really starting to fall into place though and I expect I'll have my "killer demo" by the end of the year. The problem is with this really is that I can't point to something and say "it's like that" because it doesn't exist and no one else is trying to do anything similar as far as I can tell.
Ok, now I understand. There was a European consortium working on this a few years ago for product development in the automotive and aviation industry. I think Barco had the lead on that.
I'll do some digging on the euro consortium, it wouldn't happen to be AliceVision would it? That's more a euro university group, but I could see them being related.
Privacy issues aside for a moment, this looks really good and is something I’d be interested in. I can imagine this being helpful for folks with ADHD or memory issues. I have found meeting transcription tools to be absolutely invaluable at work so to expand that to my wider life makes good sense.
That said I have to admit the monthly subscription price ($19-$29/mo) seems a little higher than I’d like for a wearable device, especially as the premium features appear functionally like a thin wrapper around an LLM and some search tools.
Indeed. I’d love that with glasses as well - having perfect photographic memory of everything I see or hear, augmented by an AI that can correlate events, would be a real superpower.
OTOH, I also watched enough Black Mirror to imagine the societal consequences of that.
And yes, the subscription model indicates they’ll know a lot more about you than the ad copy discloses.
If someone does one that’s fully local with absolutely no strings attached, I’d seriously consider it.
That pricing is for their other existing product/app that records and exposes digital activity locally in what looks like a fairly nice UI: https://www.rewind.ai/use-case/engineering
Very interesting company IMO although I haven't given it a shot.
It's incredible how every single endorsement on the homepage just reinforces the bad vibes on this thing. Sam Altman, Marc Andreessen, Alexis Ohanian, David Lieb. Genuinely expected to see an unironic Peter Thiel, Alex Karp, or John Riccitiello endorsement.
It seems only a few learned a thing with Google Glass[0]. These devices are way ahead of our time. We're currently in a time and space where implementing these ideas only serve to tighten the elite's control over people.
More specifically, devices like this are cool when not widely known or adopted.
As for "elite control", "elite" ain't controlling shit here - they're just peddling things to general population and raking it in. Outside of misplaced paranoia about government (misplaced in that spooks don't wait for surveillance equipment to be available off-the-shelf), people aren't worried about elites, they're worried about each other - randos on the streets being creeped out by other randos on the street. Society doesn't have the levels of kindness and trust to tolerate these kinds of technologies. Which is a shame, given the positives.
(Of course those technologies being developed by advertising companies isn't a good thing, but notice how hardly anyone talks about that. People are too creeped out by each other, and not enough by advertisers.)
I'm curious how this works out in the privacy of people's homes. A lot of couples wind up arguing about who really said what. I wonder how it works out for them when someone produces the recording that settles it for good.
While we would hope that it settles the argument, it doesn't address the emotions. And therefore I suspect that it would just make people's arguments worse.
Yeah, throwing a recording in their face seems like a surefire way to escalate that argument rather than resolving it. It's more about trust and mutual respect & compromise and forgiveness... if you start keeping score and playing gotchas, that only builds resentment.
>I wonder how it works out for them when someone produces the recording that settles it for good.
Portable microcassette records have been around for many decades (and hitting record on a phone to capture audio is something anyone can do without taking the phone out of the pocket).
There's a reason you're wondering about how well this idea works.
>> A lot of couples wind up arguing about who really said what.
I try to get as much relationship decision making as possible done via SMS text or email.
And the moment there is a discussion/argument that turn to “what you said/what I said”, I say I’m only willing to talk about tangible facts, not conversation recollections.
You need this product to extend your Twitter inquisitor zeal to the offline environment. /s
No joke, I remember at uni muslim girls who were obliged by their family to have their phone on permanently when not in a strict controlled group. The latter love these things.
I hope this is just prohibited in the EU :-) If I see someone recording me(whether they did or not), I'll use my legal insurance to sue them into oblivion.
There's an imaginary boundary of when technology starts to hurt humanity, and this clearly crosses it in so many ways.
I can't believe the people behind this haven't thought about the negative sides, even if they deliver what they said in terms of privacy, it's so wrong in so many levels. Did they use ChatGPT to generate that silly idea?
This is when you know AI hype has hit cryptocurrency level, where stupid ideas start popping all the time and supposedly raising money.
It's time to grab popcorn and see VCs fund those stupid ideas and see them wasting their money.
This device is very similar to what I've pursued for a short time.
Almost every country (with a notable exception of several US states) allow to record conversations for personal use. You just cannot distribute them to third-parties without consent.
If they take a picture of me and I know it, yes. You can't see my house on Google Maps.
I live in a civilized country with some privacy laws, which will likely just pump out more laws once silly products like this start popping in the market.
I have seen verbal fights on the streets of Germany where one side has pulled out their smartphone and started recording. The other person was shouting "don't record me", but the recording person was insisting it was because the other person was breaking some obscure law.
I did not sit to see what the police would have done with the material once they came, but are we sure they would dismiss it immediately, provided it had indeed broke some law?
Germans are really concerned about their privacy. Do you believe if that would be the norm, they wouldn't vote for politicians to completely abolish recording devices out of existence?
>Rewind Pendant is a wearable that captures what you say and hear in the real world and then transcribes, encrypts, and stores it entirely locally on your phone
So, the recordings are not stored locally (i.e. on-device), and it requires a phone to function.
So one of the interesting parts from Zuckerberg's interview was that they are expecting to add a local multi-modal ML model capable of image classification, speech to text etc to the Rayban glasses as soon as next year.
And with Humane having a similar device to this and Ive likely building one with OpenAI it looks this is going to be a heavily contested field.
I do wonder if these products are reading the room though. Because there is a demand at the moment for more privacy not less.
I love all this new exploratory hardware coming out. At $59 this seems like an easy way to try out the concept without committing, and the form factor doesn't seem fiddly.
I wonder if there's ways to get the benefits of this while avoiding privacy issues. The best way I can think of preserve others privacy in an unobtrusive way would be to detect & remove other people's voices from any recordings (private by default). The only way to override this would be if others say "its ok to record" or equivalent phrase.
That way you get meeting notes (where people consent) but not conversations in confidence (where people don't). For the in-between cases that users wanted to remember, they could repeat back what others say (e.g. other: 'that is $59' user: 'ok $59').
If this were accurate enough it seems like it'd address most of the privacy concerns, and is a pattern that can be used by all types of recording devices we'll be seeing in the coming years (and may be a good basis for legal regulation of this class of device).
Zooming out, we're early in the cycle of this kind of tech, where forgetting is seen as a problem. There's a chance we over-shoot and remember too much, which is going to introduce its own problems because tech doesn't have a sense for what things are important psychologically for us to move on from and not have access to. Much like in other domains where tech gives us too much of what we want, I worry the response will be a dismissive "just don't look at it" personal responsibility argument. But that's further down the road from where we are.
> I wonder if there's ways to get the benefits of this while avoiding privacy issues.
There is. Get informed consent from the people before you record them.
> The best way I can think of preserve others privacy in an unobtrusive way would be to detect & remove other people's voices from any recordings (private by default).
I would consider this very insufficient because I've already been recorded at that point. But it would certainly be better than nothing.
Better, in my view, is to get informed consent first. That gives people a real option to protect their privacy rather than relying on your (or the product's) good graces to be well-behaved.
Privacy issues aside, it's really cool to see a new category of device!
I've been actively searching for something that will allow me to "write" my essays while I'm driving.
I've tried voice memo apps for iOS, and the transcription stinks for everything I've tried.
Rewind's transcriptions for meetings works great, so I'm cautiously optimistic
---
Sidenote: I've been loving Rewind as a company, but I'm a bit worried about how they're approaching security.
They store fantastic metadata[1] in their sqlitedb locally, but they recently started encrypting the DB, which is great, BUT I want to be able to poke around and see EXACTLY what's being recorded.
Have you tried using some other speech recognition software to do the transcription afterward? In my experience Otter.ai and YouTube (uploaded as a private video) both work great for this.
I had one of those, though I used it as a way to automatically get a smattering of photos while doing tourist stuff or going to concerts, rather than lifelogging. It was also pretty great for no-effort time lapses. I'm still disappointed there isn't any decent successor that can meet the same basic functionality, especially battery life.
Probably lower quality than the Narrative Clip but I think somewhat smaller and lighter weight. Don't remember if it captures audio. It normally requires tapping on it to make it record a video clip or photo, but it can record time lapses as well. The Windows software for it is very clunky.
I mean, everyone already has a recording device in their pocket at all times that's more than capable of recording what you say if you filter the noise out. This device is just a microphone that's in a less noisy position along with a subscription.
I don't understand these sorts of responses. There's a world of difference between a device that could be abused to spy on people and a device that is designed and intended to do so.
Also, just because there are existing ways to do the Bad Thing doesn't mean we should accept new ways of doing the Bad Thing. Particularly when the existing ways have enough friction that they require someone to intentionally decide to spy, rather than accidentally doing so.
Lol. Wrote some code that basically does this a couple months ago. Whisper -> openai -> notion recording and classifying conversations. Never got it running on anything other than my laptop. I'd imagine anyone serious about this kind of thing would want to run everything locally. Plus this will be the simplest thing for any legit Ai company to crush in the future. https://github.com/bfpill/Gekko
This looks like a privacy nightmare, even for the wearer - aren't going going to have to continuously ask for permission for every single person that you're in recording distance of when not in a public space?
That depends on where you live. In some European countries for example it's perfectly legal to record for personal purposes without announcing it. Bear in mind the phrase "for personal purposes" is doing a lot of heavy lifting here, for example it's common that you can't use the recording in court if you haven't announced you're recording and since that would no longer be personal use.
Been thinking of something like this for a while, but with video as well. It would be cool to have a searchable transcription of things in my immediate surroundings. Might make it impossible to lose my keys...
Does the LLM also reside on the local machine? I presume there would need to be something to manage the transcription and address queries related to that transcription. It seems like the LLM would require a reasonably extensive context window to be functional. Wouldn’t this necessitate a relatively powerful computer? I am definitely not an expert on this so I could be completely off target on one of these assumptions.. genuinely curious though.
There's a short short by Ted Chiang about this concept (he also wrote Stories of Your Life which the movie Arrival is based off of). In the story, a father records his family's entire life but he realizes that he was just living vicariously through the camera instead of truly being present in the moment.
There was an early kickstarter for a watch that did something similar and I backed it, thinking it was brilliant.
One of the first times I wore it was to a meeting, someone asked about it and when I explained how it worked he refused to continue the meeting until the watch was in the other room, turned off.
I tried their Mac app for a while, but the storage space required quickly became impractical for a retention period set to any actually-useful amount of time. I wonder how they've handled that here.
I've been thinking about doing something like this myself. Lyra[0], a codec by google, supports speech @3.2kbps (0.0004 MB/s), so 3.26060*24/1024/8, or 33.75MB/day. Filtering out periods of inactivity, it's likely under 10MB a day.
I feel like my math must be wrong, because this feels crazy!!
I could imagine a few professional uses which are less creepy. Students/teachers, comedians, legal/medical dictation.
I think the real power would be if it is wired up to an appropriately accurate text to speech engine and dumped to a searchable database. I have been blown away by how good Whisper is, but that requires too much technical expertise for the population. Having that workflow in a packaged product would be worth something not otherwise met by existing recorders.
This is going to get some teacher unions in one-party consent states[1] up in arms as students start bringing similar types of (small) recording devices into the classroom.
It says this 3 times without elaborating. How does this work?