Google has been tastefully integrating AI into phones for a while now using opt-ins. The now-many-years old phone screening feature ensures the fear of spam won't lead to unintentionally missing calls from the doctor/dentist who call from unknown numbers. Screening suspicious calls instead of declining them is a massive quality of life improvement, and having an AI assistant answer your phone and do a live transcription is straight from the Jetsons.
And there is a feature now to show you how busy a customer service line is at a particular time, and to give you a text transcript with clickable buttons for the phone tree.
If an annoying phone tree is going to take 3 minutes to navigate by listening to the prompts, then it will also take 3 minutes to navigate by letting the AI listen and write little buttons for me.
I didn't really save any time.
If on the other hand it could say "other users have seen this phone tree before, so here is the whole searchable tree with no waiting", that would be a noteworthy feature.
There are probably only ~10,000 phone trees that cover the vast majority of calls in the USA.
If a phone tree is seen by many users and is identical, there is probably no privacy reason not to publish it. Even moreso if the total number of digits typed in quick succession is < 4 (ie. user hasn't entered some password to get in). A human operator could be the final check.
Also, it often prevents me from having to listen to the tree twice, if I am looking for a better option and then decide a previous one was actually the better fit.
If it's a common number for a support line it actually shows you the options before they're stated by the robot. Say for example if I call bank of america's support line, it will show the not-yet-said options as greyed out, but selectable. If the call robot accepts early inputs you can actually navigate a phone menu quite rapidly. Pair that with the estimated wait times or "hold for me" you can save yourself the headache of "actively" waiting.
i Googled that term
looks like it's another name for an IVR menu.
i googled this; looks like it means "Interactive Voice Response" menu, describing automated phone systems where the user navigates a series of options using either their phone buttons or their voice.
https://en.m.wikipedia.org/wiki/Interactive_voice_response
That's what they are called in India, at least, where I live, and where I am from, although the term would not have originated in India, since it would have been brought in by mobile phone companies.
https://en.m.wikipedia.org/wiki/Interactive_voice_response
Although I have a good amount of exposure to, and familiarity, with US terms and slang, due to older relatives up to grandfather level having visited the US for more than short periods, having visited the US myself, and having worked in US-based companies (but in India), the exposure is not complete, that's why I was not familiar with the term phone tree.
That should not be surprising to anyone.
>I'm on the east coast in case its a regional thing,
everyone does not live in the US.
see another reply by me: https://news.ycombinator.com/item?id=41251430
Probably because it's just bolting more shit on the ends of a shitty, outdated relic of a system. Like, we have two different computers, one of which is playing back a pre-recorded voice, to a computer which is transcribing that voice into text and buttons, so it can play a generated voice back to the other computer, which will then interpret that voice into text, then perform an action based on that text. The action probably being repeating that process a few more times.
This is remarkably bad. I'm remembering the last time I had to deal with Apple Support. I didn't sit on hold for hours, I told them I needed a call, told them what the issue was, gave a brief summary, and then their support person called me, I was immediately on the phone with a human, she sorted out my problem, and the entire thing took probably 10 minutes tops, including filling out the form.
It feels like these legitimate challenges to businesses to manage their customer service lines, support lines, etc. have actual solutions that:
a) don't require two different damn computers talking over a shitty low-baud connection to one another, and
b) result in a better experience for all parties involved
If they'd just be willing to spend a little goddamn money, like they actually care to provide good service, rather than just paying lip service to it while offering the same garbage system they've offered since the fucking 1970's.
Wrong. These are not challenges. The customer service lines work exactly as designed: as cheap as possible to operate, as effective as possible in preventing majority of the callers from engaging support staff and starting any kind of support process. And the magic, most people who try and fail to get support believe it's their own fault - they're not being patient enough, or persistent enough, or maybe the problem isn't that big a deal, etc.
The companies don't want you to work around the phone tree and hold time. And once most people have tools that are able to, the companies will adjust their support channels to compensate.
Yu are mistaken in assuming businesses want better experiences for everyone, including workflows that may lose them revenue (like cancelations).
The friction is intentional and part of a wider dark pattern strategy employed for "retention" - even Apple does this when one tries to get them to stop an unexpected recurring Apple Music subscription that only shows on ones credit card and nowhere else (apple.com or the Music app). The support number is not provided on amy of the support pages, and ine has to get their bank to stop the outgoing payments. Guess how I know this...
In light of this, I'll gladly take any client-side agent assistance I can get to make to reduce my cognitive burden; the "hold for me" emulates the support agent calling you as you described. , and the experience is more important than implementation details.
I could not understand all the Adobe hate before, but they made a neutral person negative with that experience.
Tastefully integrating AI involves not uploading all of my photos and speech and text to the cloud and doing who knows what with that personal data.
The majority of this data should never leave my device and that's why on-device LLMs/AI are much more tasteful.
Heck, Google is removing Location History storage from their side and making it device side only. Users in the past few weeks are getting notified to trigger the migration to device storage or all data is deleted.
And with every new phone I always power on it without an account, update, reset and only then add an account; I had my share of 'I did disable that thing'
EDIT:
[0] and they claimed what there is not enough storage for it. On a fresh phone with 128Gb on board. For the small array of ~30-100 letters. Fuckers.
I declined the first request that interrupted my task, then tried to select a photo to continue my work, but the "Yes" button for uploading recent photos appeared over top of the photo just as I clicked. Google Photos immediately began uploading my photos without my true consent.
I hate google so much for doing this.
In the world of laptops, we have a wonderful situation. I can select the hardware that I like, then wipe the whole disk and install the software I like. Which in my case is a Linux distribution and then exactly the applications I prefer.
I wish the world of phones was like that.
I mostly live in the Plasma desktop, Konsole and Firefox. And life is good.
What are you missing?
Certainly they're much more complex. More susceptible to malware, too.
But they're also much more powerful. I can get tasks done 10x as fast. They're much more open, too. They have well-defined interfaces for most things, which means I can automate tasks. Can't do that on a phone.
From a security and freedom perspective, I actually like the restrictions of the Android platform if implemented as Google intended, which means allowing you to roll your own ROM and relock the bootloader with your own keys. Android itself has among the strongest security models for a consumer platform, again if implemented as Google intended (which is why GrapheneOS only supports Pixels). You're actually not supposed to root your phone because that opens up a large attack surface.
It's inconvenient for customization, sure, but you can still wipe the phone and roll your own system. It's a matter of the workflow to do it.
… they would enable swapping button order.
… they would remove gestures.
… they would improve OS stability.
… they would add a hardware do not disturb switch.
… they would add hardware camera and microphone switches.
… they would add a headphone jack.
Their phones are still preferable to iPhones, because at least with a Pixel one owns one’s own phone.
I will likely pay a premium for the Pixel convenience when it's time to get a new phone. Will run this 6a until support ends.
But on the positive side i like the new 9 pro because of its smaller size + hardware improvements (better ultrawide, ultra sonic finger print sensor, 16GB RAM...)
Adding per app volume control is IMHO a really bad UX. It means there's a master volume and then per app volume is like a coefficient the app volume is multiplied by.
My car has its own volume control separate from the phone bluetooth volume and I hate the experience of having two volume knobs interacting with each other when all I want to do is to increase or decrese volume
I really prefer it for having different things set at different volumes. Like presets.
I'm surprised Pixels don't; small things like this area each QoL wins.
Samsung/OneUI has a good few addons like this. Ability for multiple apps to play sound is good too (same addon) and the ability to output different apps from different sound devices at the same time (built in)
I have the Pixel "a" devices, and hope they can stay at that price point.
> per-app volume mixer
Why do you don't like it?
And there are other use cases like running your own LLM offline or using it with home assistant.