How much longer until the AI bubbles pops? I’m tired of this.
It’s when the coffers of Microsoft, Amazon, Meta and investment banks dry up. All of them are losing billions every month but it’s all driven by fewer than 10 companies. Nvidia is lapping up the money of course, but once the AI companies stop buying GPUs on crazy numbers it’s going to be a rocky ride down.
Is it like crypto where cpus were good and then gpus and then FPGAs then ASICs? Or is this different?
I think it’s different. The fundamental operation of all these models is multiplying big matrices of numbers together. GPUs are already optimised for this. Crypto was trying to make the algorithm fit the GPU rather than it being a natural fit.
With FPGAs you take a 10x loss in clock speed but can have precisely the algorithm you want. ASICs then give you the clock speed back.
GPUs are already ASICS that implement the ideal operation for ML/AI, so FPGAs would be a backwards step.
Thank you for the explanation!
If bitnet or some other technical innovation pans out? Straight to ASICs, yeah.
Future smartphone will probably be pretty good at running them.
It’s probably different. The crypto bubble couldn’t actually do much in the field of useful things.
Now, I’m saying that with a HUGE grain of salt, but there are decent application with LLM (let’s not call that AI). Unfortunately, these usages are not really in the sight of any business putting tons of money into their “AI” offers.
I kinda hope we’ll get better LLM hardware to operate privately, using ethically sourced models, because some stuff is really neat. But that’s not the push they’re going for for now. Fortunately, we can already sort of do that, although the source of many publicly available models is currently… not that great.
There’s absolutely a push for specialized hardware, look up that company called Groq !
Yes, but at this point most specialized hardware only really work for inference. Most players are training on NVIDIA GPUs, with the primary exception of Google who has their own TPUs, but even these have limitations compared to GPUs (certain kinds of memory accesses are intractably slow, making them unable to work well for methods like instant NGP).
GPUs are already quite good, especially with things like tensor cores.
Wildly different, though similar in that ASIC was tuned to specific crypto tasks, everyones making custom silicon for neural nets and such.
I wouldn’t plan on it going away. Apple put optimized neural net chips in their last phone. Same with Samsung.
✨
We’re still in the “IT’S GETTING BILLIONS IN INVESTMENTS” part. Can’t wait for this to run out too.
Here’s the thing, it kind of already has, the new AI push is related to smaller projects and AI agents like Claude Code and GitHub copilot integration. MCP’s are also starting to pick up some steam as a way to refine prompt engineering. The basic AI “bubble” popped already, what we’re seeing now is an odd arms race of smaller AI projects thanks to companies like Deepseek pushing the AI hosting costs so low that anyone can reasonably host and tweak their own LLMs without costing a fortune. It’s really an interesting thing to watch, but honestly I don’t think we’re going to see the major gains that the tech industry is trying to push anytime soon. Take any claims of AGI and OpenAI “breakthroughs” with a mountain of salt, because they will do anything to keep the hype up and drive up their stock prices. Sam Altman is a con man and nothing more, don’t believe what he says.
depends on what and with whom. based on my current jobs with smaller companies and start ups? soon. they can’t afford the tech debt they’ve brought onto themselves. big companies? who knows.
Time to face the facts, this utter shit is here to stay, just like every other bit of enshitification we get exposed to.
as long as certain jobs and tasks can be done easier, and searches can be done faster, its gonna stay. not a fad like nft. the bubble here is the energy and water consumption part.
as long as certain jobs and tasks can be done easier, and searches can be done faster
I’m still waiting for somebody to prove any of these statements are true. And I say that as somebody working in a company that demands that several employees use AI - all I see is that they now take extra time manually fixing whatever bad output the LLM produced, and slowly losing their ability to communicate without first consulting ChatGPT, which is both slow and concerning.
The worst part is that once again, proton is trying to convince its users that it’s more secure than it really is. You have to wonder what else they are lying or deceiving about.
Mullvad FTW
Yes, indeed. Even so, just because there is a workaround, we should not ignore the issue (governments descending into fascism).
Very true
Sauce?
Zero-access encryption
Your chats are stored using our battle-tested zero-access encryption, so even we can’t read them, similar to other Proton services such as Proton Mail, Proton Drive, and Proton Pass.
from protons own website.
And why this is not true is explained in the article from the main post as well as easily figured out with a little common sense (AI can’t respond to messages it can’t understand, so the AI must decrypt them).
They actually don’t explain it in the article. The author doesn’t seem to understand why there is a claim of e2e chat history, and zero-access for chats. The point of zero access is trust. You need to trust the provider to do it, because it’s not cryptographically veritable. Upstream there is no encryption, and zero-access means providing the service (usually, unencrypted), then encrypting and discarding the plaintext.
Of course the model needs to have access to the context in plaintext, exactly like proton has access to emails sent to non-PGP addresses. What they can do is encrypt the chat histories, because these don’t need active processing, and encrypt on the fly the communication between the model (which needs plaintext access) and the client. The same is what happens with scribe.
I personally can’t stand LLMs, I am waiting eagerly for this bubble to collapse, but this article is essentially a nothing burger.
You understand that. I understand that. But try to read it from the point of view of an average user that knows next to nothing about cyber security and LLMs. It sounds like it’s e2ee that proton mail and drive are famous for. To us, that’s obviously impossible but most people will interpret that marketing this way.
It’s intentional deception, using technical terms to confuse nontechnical customers.
How would you explain it in a way that is both nontechnical, accurate and differentiates yourself from all the other companies that are not doing something even remotely similar? I am asking genuinely because from the perspective of a user that decided to trust the company, zero-access is functionally much closer to e2ee than it is to “regular services”, which is the alternative.
The easiest is to explain the consequence.
We can’t access your chat history retroactively, but we can start wiretapping your future chats.
If that is too honest for you, then just explain the data is encrypted after the LLM reads them instead of using technical terms like zero access.
This I can agree on. They would have been better served and made it clearer to their users by clarifying that it is not ‘zero trust’ and not e2ee. At the end of the day, once the masses start trusting a company they stop digging deep, just read the first couple of paragraphs of the details, if at all, but some of us are always digging to make sure we can find the weakest links in our security as well as our privacy to try and strengthen them. So yeah, pretty stupid of them.
First of all…
Why does an email service need a chatbot, even for business? Is it an enhanced search over your emails or something? Like, what does it do that any old chatbot wouldn’t?
EDIT: Apparently nothing. It’s just a generic Open Web UI frontend with Proton branding, a no-logs (but not E2E) promise, and kinda old 12B-32B class models, possibly finetuned on Proton documentation (or maybe just a branded system prompt). But they don’t use any kind of RAG as far as I can tell.
There are about a bajillion of these, and one could host the same thing inside docker in like 10 minutes.
…On the other hand, it has no access to email I think?
Why does an email service need a chatbot, even for business?
they are not only an email service, for quite some time now
There are about a bajillion of these, and one could host the same thing inside docker in like 10 minutes.
sure, with a thousand or two dollars worth of equipment and then computer knowledge. Anyone could do it really. but even if not, why don’t they just rawdog deepseek? I don’t get it either
…On the other hand, it has no access to email I think?
that’s right. you can upload files though, or select some from your proton drive, and can do web search.
sure, with a thousand or two dollars worth of equipment and then computer knowledge. Anyone could do it really. but even if not, why don’t they just rawdog deepseek? I don’t get it either
What I mean is there are about 1000 different places to get 32B class models via Open Web UI with privacy guarantees.
With mail, vpn, (and some of their other services?) they have a great software stack and cross integration to differentiate them, but this is literally a carbon copy of any Open Web UI service… There is nothing different other than the color scheme and system prompt.
I’m not trying to sound condescending, but it really feels like a cloned “me too,” with the only value being the Proton brand and customer trust.
I’m just saying Andy sucking up to Trump is a red flag. I’m cancelling in 2026 🫠
What are you considering as alternatives?
OK, so I just checked the page:
Looks like a generic Open Web UI instance, much like Qwen’s: https://openwebui.com/
Based on this support page, they are using open models and possibly finetuning them:
https://proton.me/support/lumo-privacy
The models we’re using currently are Nemo, OpenHands 32B, OLMO 2 32B, and Mistral Small 3
But this information is hard to find, and they aren’t particularly smart models, even for 32B-class ones.
Still… the author is incorrect, they specify how long requests are kept:
When you chat with Lumo, your questions are sent to our servers using TLS encryption. After Lumo processes your query and generates a response, the data is erased. The only record of the conversation is on your device if you’re using a Free or Plus plan. If you’re using Lumo as a Guest, your conversation is erased at the end of each session. Our no-logs policy ensures wekeep no logs of what you ask, or what Lumo replies. Your chats can’t be seen, shared, or used to profile you.
But it also mentions that, as is a necessity now, they are decrypted on the GPU servers for processing. Theoretically they could hack the input/output layers and the tokenizer into a pseudo E2E encryption scheme, but I haven’t heard of anyone doing this yet… And it would probably be incompatible with their serving framework (likely vllm) without some crack CUDA and Rust engineers (as you’d need to scramble the text and tokenize/detokenize it uniquely for scrambled LLM outer layers for each request).
They are right about one thing: Proton all but advertise Luma as E2E when that is a lie. Per its usual protocol, Open Web UI will send the chat history for that particular chat to the server for each requests, where it is decoded and tokenized. If the GPU server were to be hacked, it could absolutely be logged and intercepted.
Any business putting “privacy first” thing that works only on their server, and requires full access to plaintext data to operate, should be seen as lying.
I’ve been annoyed by proton for a long while; they do (did?) provide a seemingly adequate service, but claims like “your mails are safe” when they obviously had to have them in plaintext on their server, even if only for compatibility with current standards, kept me away from them.
they obviously had to have them in plaintext on their server, even if only for compatibility with current standards
I don’t think that’s obvious at all. On the contrary, that’s a pretty bold claim to make, do you have any evidence that they’re doing this?
Incoming Emails that aren’t from proton, or PGP encrypted (which are like 99% of emails), arrives at Proton Servers via TLS which they decrypt and then have the full plaintext. This is not some conspiracy, this is just how email works.
Now, Proton and various other “encrypted email” services then take that plaintext and encypt it with your public key, then store the ciphertext on their servers, and then they’re supposed to discard the plaintext, so that in case of a future court order, they wouldn’t have the plaintext anymore.
But you can’t be certain if they are lying, since they do necessarily have to have access to the plaintext for email to function. So “we can’t read your emails” comes with a huge asterisk, it onlu applies to those sent between Proton accounts or other PGP encrypted emails, your average bank statement and tax forms are all accessible by Proton (you’re only relying on their promise to not read it).
Ok yeah thats a far cry from Proton actually “Having your unencrypted emails on their servers” as if they’re not encrypted at rest.
There’s the standard layer of trust you need to have in a third party when you’re not self hosting. Proton has proven so far that they do in fact encrypt your emails and haven’t given any up to authorities when ordered to so I’m not sure where the issue is. I thought they were caught not encrypting them or something.
Ok yeah thats a far cry from Proton actually “Having your unencrypted emails on their servers” as if they’re not encrypted at rest.
See my other reply. There is no way to retrieve your mail using IMAP on a regular client if they’re encrypted on the server. And Gmail can retrieve your mails from proton using IMAP. It’s even in their own (proton’s) documentation.
There is no way to retrieve your mail using IMAP on a regular client if they’re encrypted on the server.
That is probably why you can’t retrieve your emails using IMAP from a regular client.
And Gmail can retrieve your mails from proton using IMAP. It’s even in their own (proton’s) documentation.
I don’t think it can. Where in the documentation did you find that?
And Gmail can retrieve your mails from proton using IMAP. It’s even in their own (proton’s) documentation.I don’t think it can. Where in the documentation did you find that?
An online search brought me here : https://www.getmailbird.com/setup/en/access-protonmail-com-via-imap-smtp which did looks like a documentation page about how to do exactly that. Obviously, it has nothing to do with them, and the actual details makes no sense the lower you get in the page. I’ve been had :)
They still can see most mails transit from their service in plaintext in both directions, though, which remain a privacy issue, but it has more to do with email protocols than anything.
You’re right that they can see the emails in transit if you’re not using encryption, but they never said they can’t. They are as secure as they can possibly be, and are honest about what’s secure and what’s not. I would leave Protonmail at the first sniff of trouble but I just haven’t seen anything that concerning.
Agreed.
Really, if someone wants to use an LLM, the right place to run it is in a sandbox locally on your own computer
Anything else is just a stupid architecture. You don’t run your Second Brain on Someone Else’s Computer
Now, Proton and various other “encrypted email” services then take that plaintext and encypt it with your public key, then store the ciphertext on their servers, and then they’re supposed to discard the plaintext, so that in case of a future court order, they wouldn’t have the plaintext anymore.
You would not be able to retrieve your mails using IMAP from a regular mail client if they were doing that. You can even retrieve them from Gmail, which is unlikely to support any kind of “bring your own private key to decrypt mails from IMAP”.
Yes. They support IMAP. Which means, IMAP client can read your mails from the server. IMAP protocol does not support encryption, so any mail that does not add another layer of encryption (like GPG with encryption) implies that your mail is available in plaintext through IMAP, and as such, on the server.
If that’s not enough, when you send a mail to a third party that just use plain, old regular mail, it is sent from their (proton’s) SMTP server, in plaintext. Again, unless you add a layer of encryption (assuming the recipient understands it, too), it’s plaintext. On the servers.
Receiving is the same; if someone sends a mail to your proton address, is shows up in full plaintext on their SMTP server. Whatever they do after that (and we’ve established it’s not client-controlled encryption), they have access to it.
In the case of GPG with encryption (not only for signature), then the message is encrypted everywhere (assuming your “sent” folder is configured properly). But that requires both you and the other party to support that, which have nothing to do with proton; you could as well do that over gmail.
So, no, not a bold claim. The very basic of how emails standards works requires it.
Now, I’m not saying that Proton have nefarious plans or anything. It is very possible that they act in good faith when they say they “don’t snoop”, and maybe they even have some proper monitoring so that admin have a somewhat hard time to check in the data without leaving a trace, but it’s 100% in clear up there as long as you’re not adding your own layer of encryption on top of it, and as such, you, as the user, have to be aware of that. It might be fully encrypted at rest to prevent a third party from fetching a drive and getting data, logs might be excessively scrubbed to remove all trace of from/to addresses (something very common in logs, for maintenance purpose), they might have built-in encryption in their own clients that implement gpg or anything between their users, and they might even do it properly with full client-side controlled keypairs, but the mail content? Have to be available, or the service could not operate.
Protonmail does not support IMAP, what they have is a program called Proton Bridge that locally decrypts you email then you can set it up so that your IMAP client then reads from Proton Bridge, giving you a seamless experience with one email client having access to all your email accounts.
They support IMAP. Which means, IMAP client can read your mails from the server.
Proton mail does not support IMAP. Because your emails are encrypted on the server.
Again, unless you add a layer of encryption (assuming the recipient understands it, too), it’s plaintext. On the servers.
Protonmail doesn’t claim that non-protonmail email is end to end encrypted. Any emails sent to a regular email without third party encryption will be plain text through the SMTP server, but they don’t store it. So in this case they are still not storing your emails in plaintext. Your recipient will, but that’s out of Protonmail’s control.
shows up in full plaintext on their SMTP server. Whatever they do after that (and we’ve established it’s not client-controlled encryption), they have access to it.
You’ve not established that at all. Protonmail stores that message with client side encryption and they have no access to it. Nothing you’ve brought up here suggests that anything is stored in plaintext on Protonmail servers.
I’ll just repost the same message here, for completion sake.
Well, I’ve been had. There is no IMAP support indeed, during my quick lookup around it, I ended up on a website that does look a lot like a real documentation that claim it does. My bad.
The point about sending and receiving messages in cleartext stands, as SMTP works that way, but at rest it is possible they’re keeping them encrypted.
Well, I’ve been had. There is no IMAP support indeed, during my quick lookup around it, I ended up on a website that does look a lot like a real documentation that claim it does. My bad.
The point about sending and receiving messages in cleartext stands, as SMTP works that way, but at rest it is possible they’re keeping them encrypted.
There’s some good discussion about the security in the comments, so I’m just going to say that Lumo’s Android app required the Play Store and GPlay Services. I uninstalled.
It’s also quite censored. I gave Proton’s cute chatbot a chance, but I’m not impressed.
Okay but are any AI chatbots really open source? Isn’t half the headache with LLMs the fact that there comes a point where it’s basically impossible for even the authors to decode the tangled madness of their machine learning?
Yeah but you don’t open source the LLM, you open source the training code and the weights and the specs/architecture
what do you think an LLM is? once you’ve opened the weights, IMO it’s pretty open. Once they open the training data, that’s pretty damn open. What do you want a gitian reproducible build?
Yes, several are fully open source. I like Mistral
I see I’m not the only one who is sceptical of that E2E “encryption”
deleted by creator
This was it for me, cancelled my account. Fuck this Andy moron
Well, I’m keeping mine. I’m actually very happy with it. This article is full slop, with loads of disinformation, and an evident lack of research. It looks like it was made with some Ai bullshit and the writer didn’t even check what that thing vomited.
It was Snowball! He wrote the article! Must have been!
+1, it appears they didn’t check the support page:
deleted by creator
It can’t be that stupid, you must be prompting it wrong
Eat shit
Edit: is that a tag or something for the website? I still don’t like the sentiment (or the chatbot) but if it’s not something that came from Proton then I take back some of my vitriol
Who Proton??? Nooo come on… who could ever seen this coming? 🐸🍲
