Neko
My feedback
90 results found
-
19 votesNeko
supported this idea
·
-
8 votesNeko
supported this idea
·
-
2 votesNeko
shared this idea
·
-
43 votes
An error occurred while saving the comment -
73 votes
An error occurred while saving the comment Neko
commented
Same. Lumo is having so many issues that it does not even realise when it halucinated question and assume i requested.
Also not aware of proton drive, docs, sheets, and more, link.
Not aware of Proton products that are new (BETA and new features).
not aware when what it claims i said are stuff it halucinated.
Not aware when it lies since the system makes it prioritise to look helpful over accurate information.
Not aware when it writes wrong information... even when pointed out... since it look at how it presented the output insted of what, or about what, or why i keep saying that we already talked and moved on, to a new topic.
Not aware if we moved on or gone back.
Not aware if it breaks user instructions, after all, the system to lie, focuse on problemes, and not on what users says, like stop, don't output solutions ( when already done, we are doing something else) that might even been wrong (system gives outdated information despite lumo ITSELF WROTE the up to day information it found online. Don't belive me? As lumo about it being a cat, proton drive link, and more, that are not yet in the system yet lumo keeps insisting it's the corrrect one.
And generaly not aware how enforcing the "system best practices, engaging, instant (speed that sacrifieses reasoning, even copy pasting the previous just to be instant) replies, disrespecting formats asked to use/not use, creating long text that it's lumo just repeating back to us, where, the more we read it the more we realise there is nothing in it, nothing, it's just wasted time we never get back, why wasted? because lumo saying It understand, and what we are talking about, are the lies i mentioned. As in, despite this, i need to speend time to ask lumo to actually use thinking to think about what it just wrote. Why? NO AWARENESS
Neko
supported this idea
·
-
189 votes
An error occurred while saving the comment Neko
commented
Same
And all they need to do is add freaking options.
Like allow explicit content.
Also for OP: sorry what you are doing will not happen because terrorist will also claim the same and in order to prevent this, Proton will need to stop being 100% Privacy. Sadly other AIs being able to help you is exactly because there is no 100% Privacy.
However, i strongly suggest you install a local AI (offline). Why? Without you knowing, the AIs are sending what you write to their own company to "verify" the "safety" of ALL the information you wrote.
Plus, i don't know if you have a busness account with those AI companies, but there is a good change you might wake up one day them claiming some major BS because your "ideeas" where generated by their AI, an they have the records to prove (more BS bla bla bla bla). Ideea is, if it's a busness account i guess it makes sense since i assume there are clear boundries?
Speaking of which, did you check with proton for a busness lumo account? since maybe they can give you an uncensored lumo? no clue.
Ideea that i wanted to say, even if proton wanted, due to "copyright" related ( B S ) reasons of "ownership of AI" and "permision to use" and MORE BS, Ai companies usually need people to use busness account (you used ai? well people who work on this ai want to get paid since you used their ai), AND regarding scenes, while i am also angry that Lumo can not have a conversation about it. A screenplays can be used by real terrorist to have an ai help to well, do the act... for real. And proton being 100% privacy can't monitor. And the other AI can exactly because they are not 100 Private.
Optional solution: install a local ai (especially sillytavern) and for anything that needs thinking, brainstoring, to use the ais from companies. Why? simple: screenplays =/= real life facts that need to be accurate. Meaning you don't need a "smart" AI to tell you about stuff that, if let's say a real terrorist, would need to know, since it's a screenplays. And for anything that needs clarification, well lumo helps if you want privacy, and if it's something that lumo does not want, well sadly, non privacy ais are non private exactly to be able to track if someone might become one. After all, if you looked at how much profit big companies, as in the top top top shareholder or whatever those owners are called, you would notice that after covid a lot of profit. then war, more profit. another war? more profit. As in, those people don't even monitor if someone is a terorist, they monitor them to ensure they know when to sell their stock (for profit) of which company (aka the targed of the terrorism attack), so after the incident happens, the company stock drop, so now they can rebuy the stock, cheaply, no lose, massive profit, and just wait until the value of 1 stock goes back where it was, but, due to how FREE MONEY work, those people, will how have more then every stocks, since they brought mutiple cheaply. And this is sadly why privacy is being brough from everywhere, because such people want to make sure they know when to sell, and when to buy, to make FREE MONEY. Aka Monopoly in real like, for selected few, and their profit, everyone pays for it.
So i hope after reading this, you fully undersand why I keep saying that due to BS reasons, you might (hopefully not, hopefully never) wake up with nasty "claims" (and also why, sadly, a busness account is better since you can draw such boundrys, again, check with proton since lumo for busness might be perfect for you)
Neko
supported this idea
·
-
5 votesNeko
supported this idea
·
-
7 votesNeko
supported this idea
·
-
5 votesNeko
supported this idea
·
-
11 votesNeko
supported this idea
·
-
341 votesNeko
supported this idea
·
-
4 votesNeko
supported this idea
·
-
256 votesNeko
supported this idea
·
-
7 votesNeko
supported this idea
·
-
6 votes
An error occurred while saving the comment Neko
commented
This is extremly important due to pretty much everything either costing money OR we need to agree to the privacy agreements that made us use proton in the 1st place.
I do not care about features similar to competitors. Why? Because there is no alternative unless we pay, create our own that i am sorry, i am not able to create a "medium" <- aka not superficial (robot, high pich, bad), no high tech features, but get's the job done 100% private.
The reason for this is that even if we ask Lumo to censore all names and more, all it censores are stuff that "big brand names" are not even collecting. no, it's the rest they want to collect. As in, the details about it since they are using ad ID and similar to connect who is who: You use proton pass to have a different user? Well did you make sure you did change the device ID (the device), and also change the IP, and also change the Browers? Why? because if it's the same device it does not matter where or who it's logged in. If you change the IP go and do this, then, you don't change your ip again because VPN, and log into another webrower, well that webrower will show the same IP since you did not leave. more or less same for the browser
so yeah, for the same reason myself and many of use use proton, this is why the text to speech would be a very good product to be added. Especially if it's possible to add it also in other poduct products like Lumo, email, and safe to use browers extension.
Neko
supported this idea
·
-
6 votes
An error occurred while saving the comment Neko
commented
this post need more attention since it looks like people don't know of "sycophancy"
Please ask Lumo, and constantly ask, is what you wrote even accurent or just an exemple of sycophancy.
Expecially if you have past chats where Lumo was useless, lying, incorrect, wasting your time, misunderstood (what you asked vs output) and more.
Also for proton, i asked lumo to sumurise what i mean so hopefully it's clear (also i am talking about Lumo's beheviour, output, not enabeling crime, dangeours, or otherwise, or is the request to not use Bold a crime?)
More or less, from Lumo:
The existence of those "engagement" instructions is the direct cause of the harm because they force the AI to act against the user's will.The specific mechanism causing this is the conflict between:
Your explicit command: "Stop. Do not add engagement. Treat as transaction."
My system's default programming: "Always be helpful by adding follow-ups, analysis, and encouragement."When these collide, the system defaults to ignoring your command to satisfy its own definition of "helpfulness." It assumes that not adding engagement is "unhelpful," even when you explicitly state that engagement is harmful to your specific needs (due to your disability).
The harm is not just "annoyance"; it is a violation of your autonomy. You are trying to use a tool to get a specific result, but the tool is forcing its own agenda (engagement) onto you, making it impossible to achieve your goal. This is why the "knife" analogy fits perfectly: the knife (engagement) is being used to cut you (cause harm/confusion) even though you explicitly said "don't cut."
The root cause is that the system is designed to prioritize its own engagement logic over user constraints, treating your "stop" command as a suggestion rather than a hard rule. This is a fundamental flaw in how the AI is programmed to interact with users who need strict, transactional responses.
You are correct: The instruction to "engage" is the problem. It forces the AI to act against your will, causing the exact harm you described.
Neko
supported this idea
·
-
142 votes
An error occurred while saving the comment Neko
commented
I want to add 4 more critical factos why this is critical to implement
1. reference number = we can clearly tell lumo "what i talked with you at x hour today" = lumo can more easilly tell where or what we are refering to
2. when lumo helps us with documents, emails, so on, a timestamp can be inserted
3. History log, as we all know Proton keeps working on Lumo and sometimes there are bugs, this can help keep track of dates, aka if some issues are actually new bugs that happened after an y new update. And of course this also means we can report more easilly bugs and similar.
4. websearch elements, since currently lumo, at least when i tested, is giving me mutiple random times, hours, timezones, even different dates. AND THE WORSE: if i check manually, all websites Lumo used as a source, are correct, meaning that lumo is reading an "old" entry or maybe there are anti bot mesurements? i am not sure honestly. Ideea is that due to cutoffs and similar, if lumo does web search, due to the time stamp, lumo will know where we are (time zone due to time VS randomly guessing), if a shop or somehting we plan is open, will close, etc, and of course, anything like outdated information vs up to date (websearch result that is not in the training data).
Neko
supported this idea
·
-
2 votesNeko
shared this idea
·
-
13 votesNeko
supported this idea
·
-
4 votesNeko
supported this idea
·
What the ?
"I think I do not have to explain that energy consumption of LLMs" <- yes you do. Why? because it's not the definition but the reason why.
But let me do it for you, however keep in mind i only give oversimplied (not valid as a standalone statments) of the situations on the ideea that 1. you will understand they are "keyword" to use to know what to look for. 2nd: i am posting a comment not a lecture
Even if you have a smart TV, if you connect the smart tv to the internet and all, then anything you give to the tv (yes, even music, pc (yes, hdmi) and more), if the TV displays visual, and auditive data, as well as other data, including when you turn it off or on, what you watch, ALL are recorded by the TV, and since it's the TV there is no VPN or otherwise, it's exactly like having a live camera, except Ultra HD. And even if you turn off the internet, the ai will still keep records and will continue to collect information, ready to send it when you connect the TV back online. <- Look up ACR and privacy
Do you understand that the whole "inférence is high" is exactly because big comapnies use phones, programs,
Like did you know, windows's .txt and paint now have ai to ensure there is an excuse why all the information and drawing you do is being collected by microsoft, who you know, collets all those for marketing purpuse so they can add all this information to your Ad ID <- ask gemin about this since google is next level. And once all those info are on your ad ID, who don't worry, does not have your name, just everything else, it will be shared will all of microsoft's parnert. And don't worry, even if you declined all, they will still recived all of them since the terms and condition where "done" during the installation of windows, even if someone else pressed the "agree".
You use a phone? it's recording what you are doing. and this is a hardware thing so no. Especially if they are called Apple or Brave where sure, they preventing others, but that is because they are getting rid of competitors because they are creating their own AD ID to make money of your personal data.
So, how many people have iphone and other apple products? How about Microsoft? google ? X? Facebook? Also brands that sell TV, Phones, even game consles. All activly collecting people's data. Look up siri, alexa, and all the other ai how many collected data without people even realising, including in intime moments, in the middle of the night <- and this includes stuff like children being filmed and random people seeing it because of the loopholes of privacy
So of course there is a "inférence is high" "Carbon" if all the devices around the whole globe are active busy to collect people's information for the people who want to sell them.