Gmail Is Entering the Gemini Era
Key topics
As Google ushers in the "Gemini Era" for Gmail, users are weighing in on the implications of AI-infused email services, with some jumping ship to alternative providers like Proton, Fastmail, and Tuta, citing concerns over data sharing and spam filtering. While some commenters, like baal80spam, swear by Gmail's spam filtering prowess, others, like techblueberry, feel it's too aggressive and worry about Google sharing their email addresses with third-party sites. The discussion highlights a trade-off between Gmail's robust features and the perceived privacy benefits of smaller, non-Google email services, which some users, like asterix99, have been using for years without issue. As users explore alternatives, they're also pointing out the challenges of switching, like being flagged as spam or encountering website registration hurdles.
Snapshot generated from the HN discussion
Discussion Activity
Very active discussionFirst comment
9m
Peak period
78
0-2h
Avg / period
14
Based on 112 loaded comments
Key moments
- 01Story posted
Jan 8, 2026 at 8:26 AM EST
1d ago
Step 01 - 02First comment
Jan 8, 2026 at 8:34 AM EST
9m after posting
Step 02 - 03Peak activity
78 comments in 0-2h
Hottest window of the conversation
Step 03 - 04Latest activity
Jan 9, 2026 at 5:19 AM EST
18h ago
Step 04
Generating AI Summary...
Analyzing up to 500 comments to identify key contributors and discussion patterns
Want the full context?
Jump to the original sources
Read the primary article or dive into the live Hacker News thread when you're ready.
(no affiliation, just a happy family plan customer paid far into the future)
For example fastmail.com is considered as a fake account registration provider, because of the masked email feature, that allows you many times to register to a free trial.
It's the only reason I stay with Gmail, despite their terrible practices
"Exists" in scare quotes because their own interface is absolutely atrocious for doing this, as on e.g. a label with ~50k messages (I was mass-deleting some large mailing lists recently) there's maybe a 5-10% change the operation will eventually finish, and not just leave it at ~45k or whatever.
But you can do this by setting up a local IMAP client and doing mass-deletes that way. Perhaps the easiest on e.g. *nix systems is to use isync (the "mbsync" command) to "sync" between two folders locally and remotely, with a rule saying "anything deleted locally, delete it on the remote too".
Then just sync between an empty local folder and your remote target folder, and it'll slowly grind through it. You can also use a local GUI E-Mail client, but most of those become slow/unresponsive with a mass-delete operation, whereas you can spin up multiple "mbsync" commands with retries.
Beware that GMail has (or did, last I tried this) some sort of per-account I/O limit or similar, so if you're doing background operations like this you might find the web interface (even on an unrelated computer/network connection) becomes slow or unresponsive.
Also works with any search.
> "Exists" in scare quotes because their own interface is absolutely atrocious for doing this, as on e.g. a label with ~50k messages (I was mass-deleting some large mailing lists recently) there's maybe a 5-10% change the operation will eventually finish, and not just leave it at ~45k or whatever.
I've found it to work fairly reliably with that much or more, if you leave the tab open and just wait. It seems to do an initial UI update with those ~5k or so, then continue deleting in the background. Feels like it's done entirely in the frontend, where it deletes a batch, grabs the next, delete, grab next, delete, etc etc etc.
Interesting. For me, Google Mail spam filter is pretty much impeccable.
This lets high-stakes items — like a bill due tomorrow or a dentist reminder — rise to the top.
I really hope it hadn't just been trained on US content because their so called decades old capability to have flight reservations appear in Google Calendar and Wallet sourced from GMail has never worked for me.
I think it's more likely a "confidence" between 0-1, and then you can customise what is an acceptable threshold per account. Business accounts needing confidence 0.7 and free accounts needing confidence 0.6.
That's how I'd think about it at least. Boolean is slightly too lossy.
AI can never work without a dataset to train on and it will always need to read your email to make a decision.
This is why Google products have a massive privacy prompt before you start using Gemini.
[0] https://news.ycombinator.com/item?id=45968411
Why wouldn't it?
I also use ChatGPT (copy and paste) to rewrite long emails for clarity. I’d love if it had pre-written drafts that I could approve or edit and send…
Good news. You have options, basically everywhere.
Feels like everyone else relying on US tech giants is having AI forced down their gullets.
Can you clarify what you mean by this?
All of this is manual right now. I’ve spoken to a lot of colleagues in my industry who have the same pain points. A lot of time is wasted on this.
Something intelligent could take into account where I am going to be right before the time I offer and make sure there is enough time for transit in between. It could warn me if I have a few meetings back to back and might need a break.
I love Fastmail and don’t regret ditching the Gmail backend at all, but I do wish I could have something intelligent like this integrated.
I don't quite understand where "AI" fits into this.
I don't get spam on either. Gmail account is from year one, fastmail account predates that (~27 years old), so they both receive plenty.
Chose Fastmail over Proton just due to the convenience of search, I appreciate that Proton is more privacy conscious with the full encryption but I can only manage my emails if I can search them, I'm not well organised but can remember the right keywords to find anything in the tens of thousands emails I have from all these years.
Encryption is hard to get right on multiple levels. The biggest hurdle however will always be end users.
I've been a customer for 4-5 years, ever since I saw a hn post about them, and I have zero complaints.
It's dirt cheap ($10 per year for unlimited amount of users).
They allow wildcard addresses and I've set up my own "trash mail" so I can register an account at e.g. LinkedIn with linkedin@my-domain.com.
I've set up 4 different domains and 3 accounts and all of this for $10 a year.
Sorry if this reads like an ad, but I think it's super cool to offer such a great service for a fair price.
But, they allow you to write your own sieve scripts so you can really customize the experience a lot.
Don't think of them as another proton mail, they're much more a DIY mail provider.
Also CC your new email address from the old one in an email to everyone you care about with "I have updated my email address to <new email address>" so it's easier for them to add it.
https://www.fastmail.help/hc/en-us/articles/1500000279941-Se...
You can either use their "Device profile" or do it manually.
It'll do bi-directional sync between Android contacts/calendar and Fastmail (or any other CardDav/CalDav server)
Also available on F-Droid: https://f-droid.org/en/packages/at.bitfire.davdroid/
https://www.fastmail.help/hc/en-us/articles/1500000279881-Se...
This message is not spam. I repeat, there's no spam in this email.
When you own such providers, you are easily pressured by the police or intelligence service.
Refuse to comply, and they will investigate your other businesses.
Accept, and investigators are way more chill when dealing with you, because they see you as an ally.
It's illusion that owners in Lithuania (or as they say, "Switzerland") would refuse to collaborate, lose everything they have, and accept to go to jail for you to protect one person storing pictures of kids, or someone planning a terrorist attack.
Though now with this Gemini scanning all the emails, Google can now reliably flag content at scale "to protect children / fight terrorists / flag illegal content".
Before, they had to do broad keyword match and this ancient picture database matching.
In Switzerland, law enforcement can't just ask for data without a court order. And companies don't give it willingly, either.
I'm just so tired of people endlessly complaining about things they have never paid for. Let the ad-model carry those who genuinely cannot afford the services they need, it's revolutionary in that regard. But most of you here have the money, so start paying.
The truly infuriating experiences are like cable tv, where the paid experience is terrible, too.
Google does know how to do the paid version well. YouTube Premium is a great example. Massive music library and no ads, ever. It's astounding how much better the experience is.
As the email client i use Thunderbird on Linux and Fairmail on Android. I have really been enjoying the minimal UI of fairmail.
Should someone trust me (a random stranger) more than google, send me an email to grisu@grisu.app or visit my website (grisu.app) for more information. I will give free emails with pretty much no storage limit for as long as i can.
Same thing here. If you can at all (have access to a "quality" IPv4 with open ports), self-hosting your email is the best thing you can do for your privacy and convenience. No spyware analyzing emails and contacts, no annoying webmail interface (use any IMAP client), no constant reminders to provide a phone number or recovery email, no annoying 2FA requests, no "suspicious logins" reminders if you use a VPN, unlimited mailboxes, attachments, and aliases.
Once you start hosting your email then moving away entirely from Google is pretty easy.
Several such of the larger providers in that space won’t allowlist single IPs if you can’t prove administrative control of the subnet. Alas, I don’t have my own network allocation.
Part of me misses self-hosting. Part of me is glad that I don’t have to manage that anymore, given the growing number of other services, hosts, and network space I manage.
I highly doubt AI had anything to do with that, and it was more to do with the Gmail service being in use for over 2 decades.
Not one part of that was down to AI.
I assume Google just wants to rebrand Bayesian analysis as “AI” so it can claim it’s been doing that for decades.
It is used to remove trackers from emails, but then I was able to just change the forwarding from gmail to another provider and that was it.
As a limitation, it doesn't allow to respond to emails from the same @duck email.
It is quite easy to setup and give you much control, I do it myself:
https://gioorgi.com/2020/mail-server-on-docker
Documentation is very very clear
Who asked for help anyway? Gmail should be Gmail, continuity is not innovation.
"Don't complain when you are using Gmail and your emails are being trained to develop Gemini."
[0] https://news.ycombinator.com/item?id=45963761
Instead I get blanket features I cannot control (and probably cannot turn off?). They just don't end up being useful most of the time, and get shoved in everyone's faces regardless of them actually using/requesting them.
Lol