Yeah. They did exactly that
worse? What did I miss - it was never good to start with. Alexa, Siri, Cortana, Google Assistant - all they were ever used for was set timers and play songs.
AI is the only hope to make them marginally more useful than they are.
Location-aware reminders is almost literally all I want from an assistant these days. “Remind me of x next time I’m at y, or by z time at the latest.” Is this an impossible task? I can imagine how I would code it, but maybe I’m missing something.
I think they just haven’t figured out how to monetize it, really. I agree that it’s totally codable. I could do it with tasker if it were my job, IE, I was paid to and had 8h a day to do it.
set timers
This broke for me a few months ago. It just randomly… won’t start, despite saying otherwise.
“Set a timer that goes off at 9:15 am”
*It proceeds to lecture me on the difference between an alarm and a timer, also, sets neither. *
Ah, I think the wording confuses it.
Timers are set for a duration. Alarms are set for a time. Which makes sense btw, you can’t set an egg timer to 9:15 either, you set it for, say, 21 minutes (if it’s 8:54 right now). And you don’t set your alarm clock for “in 6 hours”, you set it for 8:00.
It’s a bit arbitrary, but this is exactly where I feel models such as Gemini or ChatGPT can actually improve things, because they can more readily leap from the keyword “timer” expecting a duration to that you actually meant “alarm” from the rest of the input, you just said timer instead.
Well they are different, so why would it set one if you didn’t specify what do you mean exactly
My wife cannot set timers on our Nest Hub. It just doesn’t understand her command. I’ll say the exact same sentence right after and it’ll work. We did reset her voice profile, remove/add her back, checked all settings possible, nothing worked. Such a decent piece of hardware (speakers are actually pretty good, and the screen is decent and bright) that’s ruined by shitty software. It’s been unplugged for the last month and I didn’t even care. It’s going on Marketplace next week lol
Sometimes it can’t connect to the server (which is a completely stupid necessity).
That’s where it does the voice processing. The only processing it does on-device is the wake word and taking commands. Actually figuring out what you mean is done in The Cloud. Doing that on-device would not only make the devices significantly more expensive, but they would also rapidly become outdated.
The rest of your complaints are valid and I’ve experienced them all myself to boot.
The enshitification of Assistant is what prompted me, a few months ago, to embark on a quest to remove Google (and other cloud-based services) from my home automation setup. I’ve since swapped over to Home Assistant using Zigbee for almost everything.
I had to keep the Alexa integration going, or the other half would lose their god damned mind because apparently, that’s the only way on the entire planet to turn the light by the couch on and off.
But yeah, next up is just replacing all the light switches with zigbee-enabled ones so I can go full scary motion detection in a room thing. It’s going to be super futuristic in here, like 1998!
It was always AI, what are you talking about?
I think that’s the funniest part. Like, as far as I know, the regular Assistant uses the same approach to handling data that buzzword AI things use, a neural network. But branding (and potentially internal company politics) is weird, so they decided to kneecap Assistant in order to make Gemini look better on release.
Generative AI is different, it can generate better responses that aren’t just programmed canned responses. However I fully agree with you, they’re trying their hardest to do a bullshit rebrand with it. They could have just swapped it out with assistant and I would have been ecstatic. By rebranding it I get that same bad taste in my mouth whenever marketing elbows themselves into the conversation.
except the AI has even less features
It’s not, basic functionality has been lost, and instead of completing your requests, it initiates a search/conversation.
A few commands did work and it routed it through Assistant, but after failing my daily tasks I promptly uninstalled.
Don’t you have to install the Gemini app for all the features now? The app that specifically states it’s “experimental”
To be fair, to some degree this is exactly the use case stochastic parrots (the thing we call ‘AI’ as a buzzword) can truly excel at:
- Interpeting the bullshit we stammer out when we try to give a verbal command while totally not adhering to any reliable command structure.
- Formulating a reply that sounds like fairly natural language despite how inane the sources used might be.
So yeah, we’re finally at a real use case. Gimme! And from briefly trying it, Gemini is better at figuring things out from impresice input than Assistant was.
Make no mistake, it’s ultimately the same backend. They just swapped the processing layer between audio-to-text parsing and running inputs from them (and again on the way back). Sadly no Google Now smartness at all, we’ve lost that forever. But hey, at least this improves stuff.
