Please give an option to turn off (disable) the "AI Edit"

[ephan 24]

When I select a part of my note with hyperlink, it automatically shows "AI Edit" (AI 編集）above。

 

This AI function maybe useful for some users, but not for me. 

It's very annoying, and it's completely a distractor for me. 

Please give me an option to disable the "AI Edit". 

[Dave-in-Decatur 3,970]

I agree with you. The AI Edit button on the toolbar is enough; this popup when we select text is intrusive and aggressive. You may have a more immediate impact by emailing this to feedback@evernote.com.

[janndk 509]

And in genuine Evernote style it's also buggy and popup in contexts where it can't be used.

[Dave-in-Decatur 3,970]

Don't know about buggy, but it's definitely unpredictable. Maybe it's meant to be fun and spacy -- I mean spacious -- like the new UI. My favorite is that selecting any linked text (note link or URL) will pop up the AI Edit button--what exactly is it supposed to edit? Very artificial, not so intelligent.

[janndk 509]

In my case it pops up when selecting images. My first thought was: Wow, this is interesting. But it turned out to be just a new useless distraction on the screen.

[TURNOFFAIEDIT 4]

Please allow us to disable this annoying popup!

[Kallagh 6]

agree that both the popup and icon are distracting and unnecessary. this function should not be emphasized to this extent. it should have the same priority as other options and no more. better still allow us to turn it off entirely. some people are just spooked and annoyed by the whole phenomenon.

[Dave-in-Decatur 3,970]

On 4/2/2024 at 6:25 AM, TURNOFFAIEDIT said:

Please allow us to disable this annoying popup!

Welcome to the forums. These are basically user to user. Best way to contact Evernote and give them your opinion is to email feedback@evernote.com.

[pgielzak 6]

Yes, this option should be optional.

[Efraim Queiroz 1]

Evernote is becoming unusable. First, they included the inconvenient slash commands, with the invasive text in the note. Now, they include this AI and its even more inconvenient buttons and popups. What was for me for years and years an efficiency tool became something to reduce my productivity. And the company in its supreme arrogance doesn't care about users' opinions or experience.

[PinkElephant 8,155]

The first part describes two nuisances - real productivity killers look different. Just ignore the additional options. The grey slash advise text can be turned off.

The last sentence is unfounded and sounds like a supremely arrogant self victimization to me. If you want to make your opinion heard: feedback@evernote.com

[JRobertson 2]

The "AI edit" feature has absolutely zero value to me. That this feature exists at all means Evernote users' data can be shared with third party "AI" service providers, some of which might use that information for "training" their "intelligent" systems. That anyone using Evernote might use this faux AI chatbot "edit" feature means work produced through an Evernote workflow loses its reliability as originating with actual human talent, creating new content while appropriately referencing sources. In other words, this feature makes it unthinkable to use Evernote for professional team environments that have any text quality control needs—certainly for any professional context where sensitive intellectual property or confidential information might be at risk of being exposed or corrupted. 

For most users, I suspect genuinely informed consent regarding this feature would lead to permanent deactivation. That we have seen zero attempt at providing detailed technical information about risks inherent in this fake "AI" service, or any attempt to secure affirmative activation of the feature before it appears in the interface users see when writing notes is seriously concerning. 

Evernote should 1) reverse this, so that 2) every user must first affirmatively activate this feature before it appears anywhere as an option, and 3) do this only after users have been given the option to block / deactivate / uninstall this feature—for their own devices and for any installations operated through their teams or business accounts. 

[Jon/t 1,308]

6 hours ago, JRobertson said:

some of which might use that information for "training" their "intelligent" systems

This doesn't happen. Its in the Evernote privacy policy. Anything sent to the Ai model is not used to train it.

6 hours ago, JRobertson said:

. That anyone using Evernote might use this faux AI chatbot "edit" feature means work produced through an Evernote workflow loses its reliability as originating with actual human talent, creating new content while appropriately referencing sources

Its not a chatbot, its not creating. It just performs tasks on your note content like translating them, summarising them and checking for typos.

7 hours ago, JRobertson said:

That we have seen zero attempt at providing detailed technical information about risks inherent in this fake "AI" service

Lots of blog posts about it. Check the Evernote blog and I've written extensively about how it works on the Taming the Trunk blog.

https://evernote.com/blog/ai-edit#:~:text=Your data is always protected

https://tamingthetrunk.substack.com/p/first-look-at-evernotes-new-ai-edit

7 hours ago, JRobertson said:

or any attempt to secure affirmative activation of the feature before it appears in the interface users see when writing notes is seriously concerning

You literally have to tick a consent box before using it.

I use it most days. Very useful.

[AlbertR 540]

36 minutes ago, Jon/t said:

Anything sent to the Ai model is not used to train it.

Intelligence is based on training. How should it get better if not by learning about my habits?

I know it's very hard to implement a learning strategy that's valid only for the user who's asking for. Such a machine should take public knowledge and merge it with my strategy to tag, name and fill notes. This is what I expect to be done before answering to me. 

[Jon/t 1,308]

14 minutes ago, AlbertR said:

Intelligence is based on training. How should it get better if not by learning about my habits?

 

Its not learning. Its only performing tasks.

Ai search doesn't learn. It's really just a fancy natural language Evernote advanced search. It doesn't remember anything you did before.

Same with Ai Edit. It just performs tasks like translating, summarising and checking speeling. 

There's no leaning or remembering involved. Its just tasks being performed.

[AlbertR 540]

So they should not call it "AI" 😞. EN's "AI" seems to be only a marketing buzzword.

[Jon/t 1,308]

52 minutes ago, AlbertR said:

So they should not call it "AI" 😞. EN's "AI" seems to be only a marketing buzzword.

Oh yeah. I think Ai is a marketing buzzword for every company. ChatGPT is just fancy predictive text.

I wonder what they'll call real Ai when it arrives... maybe Dave  

[JRobertson 2]

It is encouraging that Evernote has a policy that LLM / AI service providers not use data to train their models. How exactly is that monitored AND enforced? What visibility does Evernote have into the way outside services actually handle user data? What visibility does Evernote have into precisely where and how data is stored, and for how long, and what is done, if anything to ensure zero "learning" is taking place involving confidential information or IP? Has any inspection of any kind actually taken place? Does Evernote have the ability to randomly monitor service providers? If using OpenAI services, that company is accused of violating its own nonprofit charter and stealing billions of documents from content creators to create commercial services without compensating anyone or even asking permission. How can a privacy policy guard against similar abuses toward external partners? Finally: Does Evernote provide end-to-end encryption and zero visibility into user data, including by quality control staff or devices? 

[JRobertson 2]

I don't know about others, but I have never read an Evernote blog post. I do not consider Evernote's decision to alter the interface on my devices to include an on-ramp for data to be sent to outside AI service providers to have been conducted with informed consent because someone posted a blog about it. The blog also seems more like advertising than information—promising a level of security that it is unclear Evernote can possibly provide, given the behavior of some AI service providers.

When dealing with private, confidential, and protected IP, Evernote has an absolute duty to secure affirmative fully informed consent from every individual user, BEFORE placing a button on their screen that would send their data to an outside service of any kind, let alone a provider of "generative AI" services that will take in that information, analyze and reconfigure it.

[agsteele 2,952]

1 hour ago, JRobertson said:

It is encouraging that Evernote has a policy that LLM / AI service providers not use data to train their models. How exactly is that monitored AND enforced? What visibility does Evernote have into the way outside services actually handle user data? What visibility does Evernote have into precisely where and how data is stored, and for how long, and what is done, if anything to ensure zero "learning" is taking place involving confidential information or IP? Has any inspection of any kind actually taken place? Does Evernote have the ability to randomly monitor service providers? If using OpenAI services, that company is accused of violating its own nonprofit charter and stealing billions of documents from content creators to create commercial services without compensating anyone or even asking permission. How can a privacy policy guard against similar abuses toward external partners? Finally: Does Evernote provide end-to-end encryption and zero visibility into user data, including by quality control staff or devices? 

1 hour ago, JRobertson said:

I don't know about others, but I have never read an Evernote blog post.

I think that the posts that were referenced earlier will provide an insight to the answers for some of your questions.  Since we are all users here we cannot answer anything definitively for you.  You'll need to put your questions directly to Evernote.  What we know is that if you choose not you use the AI functions within Evernote then your data is not involved in any interation with the AI provider. To enable the AI features you need to go through a double confirmation that you wish to use the service. Until you've done that, the AI options are simply links to start the opt-in process.

If you want more information you can go through a support ticket or perhaps check out some of the privacy related support content: https://evernote.com/privacy/policy

[s2sailor 2,219]

2 hours ago, JRobertson said:

When dealing with private, confidential, and protected IP, Evernote has an absolute duty to secure affirmative fully informed consent from every individual user, BEFORE placing a button on their screen that would send their data to an outside service of any kind

I'm hesitant about AI as well, but if you try to use it you will see this on the popup screen and if you are concerned about doing this accidentally just click on the remind me every time option.  I think you would really need to be asleep at the wheel to unintentionally send your data to the third-party AI vendor.

 

[JRobertson 2]

4 hours ago, s2sailor said:

I'm hesitant about AI as well, but if you try to use it you will see this on the popup screen and if you are concerned about doing this accidentally just click on the remind me every time option.  I think you would really need to be asleep at the wheel to unintentionally send your data to the third-party AI vendor.

 

I am generally in agreement with you, but from the testimony on these forums, it seems like quite a few users have AI working when they don't want it to be, and that many are entirely unaware that any content of their notes is being sent to an outside service. I also think this disclaimer does not provide any clarity about how Evernote actually prevents content from being used to train models. The disclaimer also doesn't inform users that the content they send to this outside vendor will be stored there for up to 30 days—for what purpose it is unclear.

[s2sailor 2,219]

1 minute ago, JRobertson said:

I am generally in agreement with you, but from the testimony on these forums, it seems like quite a few users have AI working when they don't want it to be

I think those posts may be about the ai edit popup window which is a related but separate issue.  Many find the popup annoying and are hoping for a disable switch.

[JRobertson 2]

5 hours ago, agsteele said:

I think that the posts that were referenced earlier will provide an insight to the answers for some of your questions.  Since we are all users here we cannot answer anything definitively for you.  You'll need to put your questions directly to Evernote.  What we know is that if you choose not you use the AI functions within Evernote then your data is not involved in any interation with the AI provider. To enable the AI features you need to go through a double confirmation that you wish to use the service. Until you've done that, the AI options are simply links to start the opt-in process.

If you want more information you can go through a support ticket or perhaps check out some of the privacy related support content: https://evernote.com/privacy/policy

Thank you for taking the time to share this. I get it. I just don't agree that it is an appropriate substitute for fully informed, affirmative user consent to the sharing of potentially confidential information with outside parties to have the highly advertised sometimes flashing, colorful AI button serve as the first step of a "double confirmation". Yes... users should be aware enough to avoid making that mistake, but I just feel there is something ethically inappropriate about a company that is PAID to provide note-taking service using the interface for that service as a Trojan horse for redirecting users' information to outside vendors.

[s2sailor 2,219]

Just now, JRobertson said:

Trojan horse for redirecting users' information to outside vendors.

Trojan horse is a bit harsh, don't you think? No one is being forced to use this, it is not a hidden trap, and as I pointed out it is quite difficult to accidentally use it.  Several steps are required.  Many are hoping to have an option in settings to disable this.  Feedback@evernote.com is a good place to send any comments you may have.  We are primarily a users forum.  This is a fairly new feature and I expect it may get tweaked some more.

[JRobertson 2]

On 4/15/2024 at 3:40 PM, s2sailor said:

Trojan horse is a bit harsh, don't you think? No one is being forced to use this, it is not a hidden trap, and as I pointed out it is quite difficult to accidentally use it.  Several steps are required.  Many are hoping to have an option in settings to disable this.  Feedback@evernote.com is a good place to send any comments you may have.  We are primarily a users forum.  This is a fairly new feature and I expect it may get tweaked some more.

I don't think it is too harsh. When a company that provides fake AI "services" asks other companies to put a button in their interface that will send user data out to their servers, it is a Trojan horse. Users are induced to use this feature without ever having asked for it. Across these forums, there are users who seem to have inadvertently activated the feature and now see it doing things without their re-authorizing it. It does not appear Evernote has any real handle on this technology or any control over what the outside providers do with user data–which is stored on that outside company's servers for up to 30 days and possibly longer (given there appears to be zero oversight). 

[s2sailor 2,219]

17 minutes ago, JRobertson said:

Across these forums, there are users who seem to have inadvertently activated the feature and now see it doing things without their re-authorizing it.

First I’ve heard of this.  How about providing a few links to those comments?

[Crayons 7]

Still, I think EN should have an option to turn off (disable) the "AI Edit"  .