Gemini can now pull context from your files, emails, and more to create and edit documents.
See full article...
See full article...
Gemini burrows deeper into Google Workspace with revamped document creation and editing
- Thread starter JournalBot
- Start date
Just what nobody wanted - spyware that can't be removed and that doesn't add anything of value to the average user.
But they've got to justify their investment into AI somehow, I guess...
I particularly like how they punish you for getting rid of Gemini by taking away features that never needed it before.
But they've got to justify their investment into AI somehow, I guess...
I particularly like how they punish you for getting rid of Gemini by taking away features that never needed it before.
Upvote
50
(53
/
-3)
Google:
Yes, yes, you technically can... for now. But the Slop Factory must grow in order to meet shareholder
Upvote
46
(47
/
-1)
"Hey, Gemini! What I really need is for you to reprogram Docs so that Docs files play nicely with Word. Can you do that? When someone sends me a Docs file and I need to do real work on it, I need it in an actual productivity application. Thanks!"
Upvote
31
(35
/
-4)
ubercurmudgeon
Ars Tribunus Militum
Time was that, when you were sent some "corporate strategy document" by middle management, you knew one of them had written it, or at least cobbled it together from various sources. Now, they needn't have, and probably won't have, even read it themselves. Just what the world needs right now, unlimited meaningless drivel, unbidden and forever.
Last edited:
Upvote
50
(50
/
0)
These are my very least favorite kinds of ars articles, the ones that simply parrot what some company says their product does with a) zero verification/testing, and b) zero comparison against their competitors. At least this was included:
How to turn off an AI feature should be prominent in any article about an AI feature.
How to turn off an AI feature should be prominent in any article about an AI feature.
Upvote
34
(38
/
-4)
ReadandShare
Wise, Aged Ars Veteran
I use Gmail, Voice, Messages and Maps. Pretty good products overall (esp. Maps), but that's enough Google already. I avoid Drive, Docs, Search and pretty much everything else Google. Thanks.
Upvote
-3
(1
/
-4)
The buzzwords of middle management combined with the verbosity of AI. Truly a golden age for us all. AI writes the slop, AI digests and summarizes the slop, and no one is happy who isn't Sam Altman or equivalent.
edit: apparently this was my 500th comment on Ars! Man I wish I had something more positive to write about for the occasion.
Last edited:
Upvote
28
(28
/
0)
For one, I am grateful to Ars for publishing fewer AI articles over the past few weeks. It's a trend I'm hoping continues into the foreseeable future.
Upvote
12
(14
/
-2)
Oh. Well.
If Google says so, then it must be okay. We can always trust The Google.
[/sarcasm font]
Upvote
16
(16
/
0)
It sure would be a shame to fall behind the industry in adding cutting-edge attack surface to your product.
Upvote
24
(24
/
0)
Upvote
16
(17
/
-1)
Goddess no! I don’t want your version of Timothy Leary trying to “help” me, while mostly using me to train it.
Two words. The second could be “you” or “off”, your choice.
Two words. The second could be “you” or “off”, your choice.
Upvote
10
(10
/
0)
The explanation is probably that they ripped out all that functioning code (package tracking and pulling calendar events, etc.) and replaced it with the Wonderful Brave New Gemini World way of doing the same things rather than maintaining a codebase with both approaches in it.
While understandable from a software engineering standpoint, I still hate it.
Upvote
13
(14
/
-1)
No, it means that it clones every byte of your data it can see for Google's consumption, retained even after you've blocked it or deleted your files from their servers.
Upvote
4
(4
/
0)
ReadandShare
Wise, Aged Ars Veteran
Actually, twins resemble each other. I think you were thinking matter vs anti-matter.
Upvote
3
(3
/
0)
Sign up for Proton Mail, services like Kagi or Ecosia, and use local FOSS alternatives, and stop handing your data over to Google to monetize. Wherever possible. Sometimes it isn't because of work--I get it. But consumers need to actually exercise some free agency and free-will for once, not expect magic to happen and things to get better with some handwavium.
The recent Gemini forcing was my final straw to dump all Google services. Didn't close my account yet--just so any emails I miss don't end up being something earth shattering.
I have a family member ticked about the Copilot myopia of Microsoft and the enshittification of Windows. They do basic basic web browsing and emails and watching media. They could use any OS from MacOS to BSD or Linux or Windows for their use. They intend to just not update Windows any more (I told them that doesn't work, MS revoked that). They refuse to learn a new OS even though their laptop keyboard is dead and machine is out of warranty. Even when "learning a new OS" amounts to a different taskbar GUI...so what do they do? Complain and sulk--instead of getting off their butts.
Upvote
1
(5
/
-4)
MilanKraft
Ars Tribunus Angusticlavius
100%.
At this point, there should be a default call-out section, added to any Ars story that mentions an invasive AI feature like this from Google / Meta, to be placed immediately after a description of what the feature does with our data and/or documents.
They should just assume the large majority of us will want it, and that it will be at least somewhat challenging to find in settings, or otherwise require a combination of settings to get the desired result (because of course it will... they won't want people to turn it off).
Upvote
8
(8
/
0)
This sounds like the dystopian horrors we were all warned about as a kid.
The last fucking thing I'd ever want is for Google to go playing around in my data. So much so that I have eliminated all the data that was in my Google account.
I still have the accounts, and still get the e-mails, but I move all of those to local folders on my computer and then deleted them from Google. I've never used any of Google's (or anyone else’s) cloud services since they've been proven to be less secure than not having one's data out there to begin with. I don't fill in preference information. I don't customize online information. I treat the Internet as I would a mall. I go there, get what I want, and leave. I don't live there. I don't work there. And above all, I don't save my data there.
I do have multiple redundant backups here at home, one of which is a rapid grab option of, well, everything. But I don't put any of it "out there" and I don't allow access from "out there" into my LAN.
It's a hybrid of rocking it like it's 1986 and 2026 at the same time.
My levels of trust in any e-service are exactly zero. While some of them are necessary evils, since the services they provide are no longer offered by the other providers I once used, and I still have to use them to be a part of modern life, And I know I will never achieve zero Internet intrusion into my privacy and data. But I do what I believe will limit the scope of that scrutiny by a vastly greater degree than the average.
What will change MY mind about this?
When the people who take our data without our explicit permission let us take their data without their explicit permission.
AKA: Never.
Upvote
5
(5
/
0)
The problem with this mindset...It assumes Google will honor your request.
Remember. Google is the Home Shopping Network of the Internet. You hand them your data, for free, and they make money selling it. Colorful 3rd grader branding and tech-bro hipsterism graphics aside...Google are a surveillance capitalism company, fundamentally. Expecting them to allow you to not be mined--fundamentally confuses your relationship with Google. You are the product, not the customer.
A similar deal to those "Do Not Track" cookie request settings in browsers. Yea, they still do. In fact more so probably.
Upvote
5
(5
/
0)
There's something that I find particularly insulting/telling about how some of the earliest AI workspace features were promoted: Summarizing documents and generating them from key points. Like a tacit admission that the makers of this AI software think that most of what you read and write is overfluffed bullshit that served no purpose.
If an AI can generate an essay of paragraphs from a short-list of key points, why the fuck do you need paragraphs? It clearly can't be for readability because the guy on the other end is being sold an AI to reduce it back down to key points. You can't be pawning off the task of analysis onto the AI because that's what you're being paid to do and paid to understand. Are you just engaging in a pointless ritual because you need to be perceived as a professional? Congratulations, you've been wasting our time, and now you're pumping out more time-wasting material of a lower quality.
These aren't tools for productivity, these are tools to maximize the unproductive work you're doing.
If an AI can generate an essay of paragraphs from a short-list of key points, why the fuck do you need paragraphs? It clearly can't be for readability because the guy on the other end is being sold an AI to reduce it back down to key points. You can't be pawning off the task of analysis onto the AI because that's what you're being paid to do and paid to understand. Are you just engaging in a pointless ritual because you need to be perceived as a professional? Congratulations, you've been wasting our time, and now you're pumping out more time-wasting material of a lower quality.
These aren't tools for productivity, these are tools to maximize the unproductive work you're doing.
Upvote
15
(15
/
0)
Boy, won't they have egg on their faces when this all implodes and they have to put the old way back in again.
There's probably some senior Amazon engineers that might be available...
\s anyone?
Upvote
3
(3
/
0)
It is insane to me how they are destroying the reputation of their own features by chasing short term metrics.
Like the design of all these AI features seem to be trying to maximize accidental activation in order to game their metrics. Great for your short term numbers, I guess. But all they are doing is turning their users against these features by making them so obnoxious with unwanted activations, cluttered UIs, and buttons practically leaping under your cursor trying to steal a click.
Not that I think these are particularly compelling features to begin with. But they are really not helping themselves by building them in a way which pushes people into the "can I turn it off" camp.
Like the design of all these AI features seem to be trying to maximize accidental activation in order to game their metrics. Great for your short term numbers, I guess. But all they are doing is turning their users against these features by making them so obnoxious with unwanted activations, cluttered UIs, and buttons practically leaping under your cursor trying to steal a click.
Not that I think these are particularly compelling features to begin with. But they are really not helping themselves by building them in a way which pushes people into the "can I turn it off" camp.
Upvote
1
(1
/
0)
There's been a made race with LLMs...to find ANY use possible for them--to justify continued VC funding.
For employers:
AI will let you fire your workers!
This has spectacularly blown up in the faces of people who use it
For people in legal trouble:
AI lawyering! AI can save you money and write your divorce for you!
Except...this is literally illegal. Businesses selling these services have been sued by State Bar associations out of existence. And lawyers using AI in their arguments have been sanctioned in courts by angry justices and magistrates
For workers:
Let AI do whatever your work is! Surely it can do it faster.
Except it screws up a lot...and when it doesn't you made yourself redundant and lose your job an AI can do.
For students:
AI can cheat on your homework for you! You don't need to actually learn anything pay us money!
Which....fair point. I don't have a counter to that one. Other than...you know...being kicked out of school for academic dishonesty and having no skills to get a job with.
Upvote
9
(9
/
0)
My career has involved a lot of reading and writing, and I'm old enough to remember when that blank page was a blue screen in WordPerfect or an actual piece of paper sitting in a typewriter.
But none of this is anything that I've ever wanted or asked for. I don't want a word processor to take a bunch of unrelated information and turn it into a document. I get the distinct impression that the people designing these products have never worked in any industry outside of software programming, since they seem to believe that writing is just about putting information together.
Pro tip: if propaganda is so useful, don't you think poli sci classes might include it in their writing classes? Oh, sorry, that's a liberal arts thing, an archaic idea out of place and time in the modern STEM era. Do carry on with your plans, I'm sure this will have an infinitesimal effect on the production and dissemination of actual knowledge and persuasion.
But none of this is anything that I've ever wanted or asked for. I don't want a word processor to take a bunch of unrelated information and turn it into a document. I get the distinct impression that the people designing these products have never worked in any industry outside of software programming, since they seem to believe that writing is just about putting information together.
Pro tip: if propaganda is so useful, don't you think poli sci classes might include it in their writing classes? Oh, sorry, that's a liberal arts thing, an archaic idea out of place and time in the modern STEM era. Do carry on with your plans, I'm sure this will have an infinitesimal effect on the production and dissemination of actual knowledge and persuasion.
Upvote
5
(5
/
0)
MilanKraft
Ars Tribunus Angusticlavius
For sure the best solution is "don't use the Google suite unless boss-man says you have to", or anything from Meta for that matter. My suggestion follows the premise that Ars should always provide that useful bit of info to people reading about the latest AI BS, so that if they've had to use one of those products for a long time, and don't want to use AI or let AI train on their data, at least turning it off wherever possible is better than nothing.
Doesn't mean Google will abide or not move goal posts in 3 months so they can begin collecting again (with or without your consent), but every article should at least plant the seed: "maybe you will want to turn this off or at least disable for now — here's how".
Upvote
3
(3
/
0)
It's been fascinating listening to several podcasts suddenly start promoting European based technical services explicitly stating it gets US orgs around the CLOUD act and allows them to take advantage of the EUs privacy protections.
I'm almost certain they smell blood in the water. I'm seriously considering some of those providers for my own organization.
Upvote
0
(0
/
0)
What pisses me off about all of that is that the LLMs are dragging the rest of AI down with them. I've been doing AI-related work for nearly three decades in geology and geophysics where it can offer some interesting insights. But now that LLMs have crapped all over the idea of AI, it is going to be a lot harder to justify their use. And I'm pretty sure the same thing is true in medicine, biology, and any other field where non-LLM AI has been used.
Upvote
6
(6
/
0)
the main point in EU (I'm French) it's not to save the world, even less to save you
it's to build an industry, computing and others, without being dependent on YOU and YOUR whims.
AT LAST, after decades of lobbying, thanks to your GREATERESTESTEST president ever, ALL the europeans political parties are convinced we need to STOP just buying your computing services and to build our others
so the point in EU is not really to refuse or deny IA or collecting data or whatever politicians love to do. It's to remove your capacity of HARM from our countries.
Don't be worried, it's mostly IMPOSSIBLE. Your military and your computing industries are SO POWERFUL you are able to threaten us and at the same time seduce us.
to remove google or microsoft from our enterprises ? it would need 20 years of hard expensive crazy work
to build an ecosystem to build phones like iPhones or android without being dependent one one (more) USA companies would ask DECADES of top priority work.
USA built its computing industry with tremendous amount of money since the 1960s years, actively. With a total and full support of the government (national and federal) without never stopping.
And you merged the computing industry in the military.
But, this is the goal in EU, slowly, piece by piece, first the most important one (payment, processors for weapons and space, system softwares.. banks, )
then, if you didn't destroy our own governments before, we will see what is possible next.
-
to me, if it was possible to have a nuclear program (civil AND MILITARY) without YOU (not ONE us-based company), I would say it would be QUITE extraordinary alone!
It's THAT much we are fucked and you are GOD-powerful.
and you are again now, these very days, prooving how much you are terrifying.
EU needs to build its own tools or we will be dominated, even more than we are today.
Upvote
-1
(2
/
-3)
But all the bros are Vibe Coding your stock options and bond purchases and Vibe Compiling your insurance premiums so are you really safe?
I, for one, confabulate our AI overlords.
Upvote
2
(2
/
0)
It's a shame that on a site dedicated to the art of technology so many see "AI" and feel the need to shout "AI slop". I'd love to see a more differentiated conversation. What ever happened with "cool, new technology!" mindset?
I can see how these tools might not be super-useable for the ocassional collaborative work, but if it's an important tool you use a lot of the day, these tools are really helpful.
I can see how these tools might not be super-useable for the ocassional collaborative work, but if it's an important tool you use a lot of the day, these tools are really helpful.
Upvote
-6
(1
/
-7)
What on Earth would lead anyone to believe that "research and writing" type of work isn't collaborative? Do you think that everyone writes things solely for computers to read?
I can see why these things look great if you're writing for computers. If you're writing down ideas that you want other people to read, the content really does matter.
Upvote
2
(2
/
0)
The focus is an occasional, don't hang yourself up on the collaborative point.
Upvote
-1
(0
/
-1)
Help us understand why we should discount the importance of these tools' clumsiness in collaborative use cases when that use is a touted feature in this new release.
Upvote
1
(1
/
0)
Yes, don't read it. I refuse to read anything that's been touched by genAI (well I'd like to anyway). If you won't use your own brain to write something, why would I use my brain to read it? I want to read your voice and thoughts, not text generated by a massive tech corporation's extractive product.
Or it's fun to think about the logical conclusion of these chatbots - you could just use genAI to read and respond to it, and then the middle management will use the same chatbot to respond to your email, and so on until the text just spirals into -
Here's the thing that nobody's saying:
One.
Word.
Sentences. Liked in Linkedin.
The whole company is just burning tokens to simulate their employees communicating with each other via a single LLM, it's hilarious
Upvote
3
(3
/
0)