Comment by vessenes | Hacker News Reader

Comment by vessenes | Hacker News Reader

Parent Full thread

Just preordered -- glasses request: there's no way to say it nicely, I have a big nose. With a high bridge. Many glasses are really tough to get properly placed; may I request you not give up on the pads w/metal adjusters that you have in the product photos? Don't let a designer take them away from you, us folks not one standard deviation from 'normal' noses will thank you.

Second, how tough is this signal processing? Will it be done on-glasses, or through an app? Third, what's the equivalent of a wake word for these? I'm reminded of the Apple Vision Pro radar to see if your fingers are clicking. I could imagine a lot of modalities where I want to just start talking and have things happen, and a lot where I don't want the glasses to pick it up. Fourth, I think a cool voice change demo would be rad -- not sure what latency is like, but if you could get it end to end under like 250ms you could probably have someone exaggerate the mouth movements and pipe out audio of them talking like someone else. Which would be another really cool demo.

Super cool. You're going to be working 80 hour weeks for the next year, I'm excited to see what you put out!

> there's no way to say it nicely, I have a big nose.

But there is!

"Érase un hombre a una nariz pegado, érase una nariz superlativa, érase una nariz sayón y escriba, érase un peje espada muy barbado" (Francisco de Quevedo, 1580-1645)

Or, as translated by Timothy Adès (https://timothyades.com/francisco-de-quevedo-1580-1645-a-una...):

"There was a nose, which had a man appended. It was a nose superb, superlative, it was a nose, a hangman and a scribe, it was a swordfish, sharp and hairy-ended"

Is that considered nice? I thought Quevedo was making fun of Góngora.

Some more of their poetic battles (in Spanish): https://lclcarmen1bac.wordpress.com/2026/05/07/quevedo-vs-go...

Yep, we have been quite set on nose pads to keep them comfortable. I will keep you updated with the designs!

The decoding is done via a companion app on your laptop or phone in the cloud, but end to end latency is still quite fast, under 500ms usually.

To begin with we don't plan to have a wake word, but rather a trigger key on your connected device that you're entering text into (pressing Alt for example), similar to Wispr Flow. We're doing some R&D into the silent speech equivalent of voice activity detection, so we have a low power basic RF detection always-on, not enough to decode speech from but enough to detect rough movement, and when it detects speech-like activity beginning to start we can trigger full sensing within tens of ms. This would likely ship as a software update a few months after shipping, though.

The voice change demo is something I've thought about actually, but we never got around to doing. Might film it soon :)

Also I note from your comments here you don’t plan on local inference - that seems like a huge self own, provided the model is less than like 2b params. I’m not saying people won’t in practice prefer your cloud version, but you have this huge uphill battle ahead of you on privacy which can be wiped at the stroke of a pen with “yep it will run on your Mac.”

You can get both markets (ease of use and technical) by allowing it. Also i would bet that you gain almost no product security locking off local inference - the hardware is the hard bit. Once someone has your hardware in hand it will not take them long to reverse and build their own processing layer; you’ll need to fight those people on patents and product quality.

Heck for a v1 just allowing local sensor data would probably be enough for the tech enthusiasts. But please don’t lock this off from the get go - it makes the sale harder and I don’t think it gives you much in return.

Agreed.

Cloud processing can be tempting as the primary revenue stream, but it entirely cuts off those users who cannot send arbitrary PII over the internet (regardless of encryption promises).

It also invites the (especially now-LLM-empowered) enthusiasts to reverse engineer it.

Instead, consider selling the glasses with >20% profit, and give local users a cheaper subscription to something else (maybe the model updates?) so they can also help keep the company sustainable.

They did say they hoped to support local models in the future

https://news.ycombinator.com/item?id=50346346

I will happily pre order once a local option is available at launch

Super cool. For a cool voice change demo we'll allow a gb200 over low latency wifi :) elevenlabs has deals with a number of professional actors and voice actors for licensed uses, seems like you could probably find an ethical and cool co-brand opportunity in there.

Have fun building this! It's worth doing.

Wait. You are taking the RF response, sending it up to the cloud to interpret, and then sending the result back? Hard pass from me, sorry. Cool tech, but I don’t want my internal thoughts anywhere near the cloud. Also, this would mean I couldn’t use something like this on the subway when the phone signal goes out intermittently, right?

The commercial reality I’m sure they’re dealing with is that the percentage of users who care enough to deal with everything required for local inference is nowhere close to large enough to make it into an initial launch.

Fair enough! We take data security and privacy very seriously and have built many safeguards to prevent even us from accessing the raw data in the cloud. At some point in the future we hope to launch local inference models too!

Respectfully, talk is cheap. Every company says they take security and privacy seriously even after their database has been breached and their customers' plaintext passwords and credit card details have become public information.

If you were actually serious about prioritizing privacy above securing an ongoing service relationship you would ship local inference as the default and only (possibly later) offer a cloud service as an optional add-on for those who want/need it.

This is the gate for me. I've put a deposit down, but need to see how this is going to be run locally before I'll actually pay for them. There is absolutely zero chance of me sending everything that goes to my "keyboard" to my system via a cloud, no matter how many assurances are provided. Maybe I can see that being necessary on mobile (maybe?) but on a laptop it is a full-stop "no" - I have lots of horsepower, this needs to work 100% locally. I might be OK with doing a cloud-assisted "training" interval, but once complete, there should be no requirement for network connectivity for production use.

Same for me. I put the deposit down but will pull out if it doesn’t come with local inference. Makes no sense for my thoughts to go to the internet. I’m ok with having slightly poor local inference quality compared to the cloud one. And one can encrypt the local inference to prevent reverse engineering the weights etc.

Echoing others on privacy and security while also concerned about service availability and reliability. Offline capability would be really meaningful for a variety of reasons.

I just want to say 500ms isn’t fast.

Yes, this is the unoptimized speed while we are in development. We are working hard to make it much faster for production!

> The decoding is done via a companion app on your laptop or phone in the cloud, but end to end latency is still quite fast, under 500ms usually.

Why don't you do this on the phone? You don't want to have your proprietary tech on the user device? Or maybe you wanted to save the recording on your cloud anyway so you can train on it later or otherwise use the data?

Is there a way to not have someone inspect me uttering profanities under my nose in the cloud? I.e., can I self host the inference?

I don’t want to talk to any device. Ever. Even if it’s just mouthing words and not speaking out loud. I understand that this can be an accessibility feature that some people might want or need to turn on, but it’s not a useful feature for the general population.

Using your mouth to communicate is low bandwidth. I can type way faster than I can speak.

Using your mouth also has a high error rate, even amongst two native speakers.

It is linear in nature, unless you record it and then you can scan back and forth. But then you’re in violation of wiretapping laws in a lot of jurisdictions.

Because it is low bandwidth and linear in nature, that means it is also high latency.

It cannot be strongly authenticated. Which means it is trivially easy to spoof.

Once everyone gets used to everything being recorded, that then leaves you with no way to validate that anything you see or hear is authentically being said by another person, unless you hear and see it for yourself, with someone standing a few feet away from you.

If you are in a situation where speech (or subvocalization) is your only choice for communication, I can see this being useful. But not otherwise.

I have the opposite problem. I have a very narrow, low bridge and glasses always sit too low. Plastic frames are a no-go because the sweet spot is like an inch below my eye line.

Adjustable nose pads and spring loaded arms are priceless. So are hybrid metal/plastic frame that offer the best of both worlds.

Reply on news.ycombinator.com