"Marquardt also promised “a new experience, powered by Gemini, to home devices like speakers, displays and TVs” with details to come “in the next few months.” "
I'm sorry Dave, I can't let you do that until you've paid your subscription...
Google has decided to silence its voice Assistant and replace it with the Gemini AI service. Assistant debuted in 2016 and offered what Google called a “conversational interface” that could “enjoy entertainment, manage everyday tasks, and get answers from Google”. Receiving assistance from the Assistant required users to …
When Gemini first came out I switched over to it and found it kept doing the chatbot thing of "answering me" instead of triggering the commands. I couldn't turn off my lights, so I turned off Gemini.
Unfortunately I haven't been able to test if they have resolved this issue, as Gemini does not work for me any more. If I set my phone's assistant to Gemini then it ignores me! I set it to flash the screen when activated, and when I say the activation phrase it flashes the screen and the microphone comes on for exactly 2 seconds - but that is all that happens!
This is a shame, because elsewhere I have found Gemini to be the best performing model. I have put Gemini on my vscode both at work (with permission) and my personal system. It gives useful suggestions more often than copilot and is actually worth replacing the default autocomplete complete with. Additionally I have been testing putting the Gemini multimodal API inside applications and it is impressive when properly integrated in to something.
I'll not comment on your own use of AI because, that's like, your opinion, man.
But please leave your "testing putting the Gemini multimodal API inside applications" in the testing phase only, to spare the rest of us that don't want AI shoved into everything for a multitude of reasons. Thanks.
I'm so tired of the enshittification of technology that's being exponentially sped up through the introduction of AI into everything. It's the reason I've switched to LineageOS, have a dumb tv, and exclusively run Linux. I don't want to wake up one morning to a critical application I need to work having its entire interface replaced with a text prompt that will attempt to do what you want for you (and do it badly).
I envision a near future where our security vendor implements AI into their systems and when provisioning a badge during onboarding it invents a non-existent facility, gives this new user admin rights, and changes the lock state of every door to unlocked except for the fifteen seconds after a user scans their badge.
But please leave your "testing putting the Gemini multimodal API inside applications" in the testing phase only, to spare the rest of us that don't want AI shoved into everything for a multitude of reasons. Thanks.
I have a form of neurodversity where I have trouble with things popping up when I’m typing. I therefore have a very great problem with Outlook etc. suggesting words after the one I’m typing at that point. I only use outlook at work and the AI on windows attempts to guess what I’m going to type next and puts it up on the screen. That gives me issues and if I shut the AI down in Task Manager when I start typing it comes back. I have to use a workaround to avoid this, one of which is to use Notepad to compose emails and then copy and paste it into the reply. I can’t delete or rename the exe because I don’t have admin privileges.to let me do that.
I am just worried one of two things will happen:
1. They will give frequent and annoying pop-ups whenever you accidentally do whatever common gesture they map Gemini to, and tapping the wrong spot will subscribe you. (My phone already asks me to activate Gemini every time I pull from one corner of the screen).
2. Tie important non-LLM functions to the subscription as well, so that your life gets harder if you don't pay. They will act like this is an upside somehow.
In many ways, assistants were promised to be like artificial intelligence, in the sense they would understand what you meant and would react to your commands. That's when the movie Her came out. The problem is that there was hardly any intelligence behind, so each separate action had to be coded manually and maintained individually by an army of developers. They were already having trouble maintaining existing actions when ChatGPT came out, and that pretty much sealed their fate. I'm not entirely sure that AI stuff like Gemini can even replace them though. It's probably going to be a while before Gemini can be asked to turn on the lights or play a song — and I wouldn't be surprised if the connections to the light and sound systems and everything else still needs to be maintained by an army of developers.
I actually quite liked Google's approach, though I never used it to things that it was easier to do manually and didn't require data harvesting. At the time, chatbots could only really worked in vertical domains, but quickly became a replacement for many tedious telephone menus "press one for technical support, two for accounts, etc.…" and this is where Google quietly made some money by selling them to industry.
LLM means, in theory, that it's easier to provide more general support and, of course, the assistants provided a treasure trove of "conversations" on which to train them.
For all the shade Apple gets, their cautious approach to AI is good for users - as they actually consider it be useful AI, rather than AI for the sake of it.
Saying that maybe this is a sign to move to everyone's roll your own AI?
——————-
Oh you mean the same Apple who after the latest update re-enabled Apple Intelligence without informing me? I’ve learned to switch off the wifi when applying updates specifically for this reason. I’ve also learned to check these settings after every update specifically for this reason
Knew they'd be in there someplace. Gf bought a "smart" Samsung tv the other week. She said setting it up was hell, the interface loaded with dark patterns from which there was no escape until you'd provided them with detailed personal details and promises that they could do anything they wanted with said details.
Then after the data grab, I presume soon they'll be after cash for it to keep working.
If anything asks if you have internet, just say no.
disable Wifi, add whatever smart stick you trust
One of our TV's the makers (Sony) used such a shit processor that many apps removed themselves, so disable wifi and I have a Roku in there - though only for Plex as we get the apps we use via an STB
The other is a newish Fire TV - I would disable all that shit, but the kids use that room more and want its features so they are logged in with our son's account, otherwise I would disable it as well and be the same as the Sony - STB or a Roku for plex
That's the problem with all "smart" TVs - a few years down the track support ends and then the apps either stop working or simply get deleted. It's a scam that's been going on for more than ten years now and I don't understand why consumers still give it any attention. Or are there a significant number of people who buy a new TV every three years?
Given their past shenanigans, I wouldn't trust a Samsung TV to actually have obeyed an instruction to disable Wi-Fi, as opposed to silently searching for a nearby open access point and spying on me.
I've always refused to have any of these devices in my house on two counts: Firstly major privacy concerns, secondly; they are really annoying. It's really hard to disable on a phone as well - no single on/off setting... There is one exception where I occasionally use Google Assistant and that is: in my car...
My previous car had Android Auto which was exceedingly disappointing: The primary function I used it for was to play my kids' random song requests whilst driving.
But it would often go like this:
ME: 'Play Baby Shark'
ASSISTANT: 'You need to set up Youtube Music'
ME: 'Ask Spotify to play baby shark'
ASSISTANT: 'Ok, asking Spotify to play Raining Blood by Slayer'
I can't remember which songs we actually involved but this exact scenario arose more than once. At least the kids found it hilarious!
Sometimes the default song would be the wrong one, so you can say 'Ask Spotify to search for the birdie song' but then you need to use the touch screen to chose which version you actually want.
My latest car is built on Android Automotive (AAOS) which is similar but different to Android Auto (Seriously, who passed these product names through Marketing!?!?) and the assistant on that is a step worse.
ME: 'Ask Spotify to play Baby Shark'
ASSISTANT: 'Sorry, I didn't understand that'
ME: 'Play baby shark'
ASSISTANT: 'You need to set up YouTube Music...'
ME: 'Open Spotify'
ASSISTANT: 'Opening Spotify'
But now the only way I can find to voice search in Spotify is to interact with the touch screen to press the search icon and then press the 'Speech to text' button. Given this is all on a touch screen it requires me to divert my attention from the road to achieve this :-(
At one point one of my children really got into a French band called 'Kids United' and would ask me to play things like 'On Ecrit Sur Les Murs' - It's safe to say that Google Assistant cannot handle non-English words when set to English. I'd hazard a guess this means that you're a bit stuffed if you drive your car onto the continent unless you want to go the whole hog and talk to your device in the local language as well.
So if Gemini can handle these tasks then it will be a step forward.
However, the whole voice interface needs to get a whole load better to be genuinely useful - For starters I do not need it to repeat back to me everything I've asked it to do. It would also help if it were less disruptive - i.e. when a song is playing I can't queue up another one without it pausing the one being played. Ideally I'd be able to just say "Play agadoo next" and it would add it to the queue silently and without interrupting Baby Shark.
Likewise I can now fairly reliably press the mic button on my steering wheel and say "navigate to Westward Ho!" but when there are multiple route options and I want to select the one which is only 5 mins slower but 20 miles shorter, if there's a way to do this verbally I haven't figured it out yet so it's back to interacting with the touch screen again (Side rant: I hate touchscreens and they are nothing short of dangerous in cars. Bring back physical buttons!).
So let's see what Gemini brings. If I can control all the bells and whistles of my car by talking to it in plain English then in spite of my general AI/LLM scepticism I might be persuaded to use it,
What will that new experience be?
The experience of being bashed to tiny pieces with a sledgehammer? Or being fried with thousands of volts up every GPIO pin in a microwave? Or a bit of 'experiential tourism' to the nearest landfill?
If only Google hadn't locked down the bootloaders on their devilish devices, one could save some e-waste and give those poor devices the experience of running free software
Oh well, off to Silicon Hell you go
'Droid and smart home kit will get a 'new experience'
The 'New' experience will be very similar to the 'Old' experience.
It will not work ... BUT only after you expend a huge amount of time trying to get it to work !!!
Personally, the 'Old' not working is just what I want and I will not try the 'New' not working if at all possible.
I do not need any 'AI' crud ANYWHERE and look forward to the subscription so I can NOT pay it !!!
:)
Countdown timers, some music, and the one he missed--> weather. We set so many timers so often we got the echo clock--and it is great.
However, I'll never understand how these are supposed to make money. MS agreed and killed-off the cortana assistant. if they want a subscription-based assistant to do those 3 things... I'll just repurpose an old tablet and have it display the weather and hook into my lan's music collection. and spend $5 on ebay for 10 digital mini timers.
I use Alexa to play Scala Radio to my dog overnight.
I use Alexa to set timers for cooking.
I use Alexa to do lights but SWMBO doesn't like it.
The only other use she gets is making animal noises when our daughter comes around.
I may be unimaginative but I don't want much else
I tried really hard to use my echoes in the way Amazon wanted. The audio never matched what I said, and the more we tried to use it the more it kept pushing sales onto us or other devious prompts to enmesh even further into the ecosystem.
Our echo shows are amazing - when they’re showing the content we want. The problem is some stuffy product manager gets new toggles and landing screens added all the time. So while we see pics that we wanna see, we suddenly get landing screens for Amazon Prime or shopping foisted on us. I spend another twenty minutes finding the new toggles, the landing screens go away for a while, and then the process repeats.
That’s why Echoes aren’t taking off. Poor response, continually having our settings overridden. Apple should be better in this respect.
int main(enter the void)
...