Login
You're viewing the functional.cafe public feed.

Replies

  • Nov 7, 2026, 11:08 AM

    Interestingly we have started to test AI infrastructures on private European providers and open weights inference. If the quality is enough and the costs not too high we are progressively move all our AI needs to these new infrastructure. Super excited by all this!

    💬 8🔄 5⭐ 23
  • Nov 7, 2026, 8:02 AM

    @gael You mean so that a version of it comes into existence that isn't based on theft, ecological destruction and amoral bias? Until that happens, being good people and embracing it are mutually exclusive.

    💬 1🔄 5⭐ 42
  • Nov 8, 2026, 12:19 PM

    @kerfuffle @gael There are many aspects of our civilization that most of us rely on which have mixed impacts. The items we all consume come from sources which could be more ethically and regeneratively managed. Here we are exchanging messages on a distributed computer system that consumes almost exactly the types of resources as AI for example. I think the search for better ways forward is much more productive than calling for outright avoidance

    💬 3🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 8, 2026, 9:54 PM

    @metaphase @kerfuffle @gael The same types of resources, but in vastly, VASTLY different quantities, and to vastly different ends!

    It's a lot easier to resist something before it becomes tied to our ability to sustain ourselves. NOW is the time to resist, not later.

    💬 1🔄 0⭐ 0
  • Nov 14, 2026, 6:19 PM

    @lykso @metaphase @kerfuffle @gael It’s also unclear that AI has any real utility outside of very skilled and specialized applications like cancer research, resolving images of distant black holes, and mechanical girlfriends giving tips on suicide.

    💬 0🔄 0⭐ 0
  • Nov 9, 2026, 8:51 AM

    @metaphase @gael@mastodon.social Equating the use of a decentralized message board, that anyone can run a node for on a consumer computer with few resources, and which serves people to choose what they want to publish to the world, to the use of AI which requires datacenters that consume more electricity and water than small town and is trained on stolen data, is a gross false equivalence. We're not resisting something because it's new or not useful, we're resisting something because it's harmful.

    💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 7, 2026, 8:46 AM

    @gael
    The only powerful effect will be to discriminate bad people, deluded people and supporters from the rest of humanity.

    💬 0🔄 0⭐ 4
  • Antinantinomian@todon.nl
    Nov 7, 2026, 9:04 AM

    @gael

    Yes, just like fission reactors were! Yikes. Good to know what I can expect from Murena and /e/OS in the future.

    You "big tech alternative" guys really know how to alienate your market audience, eh? Do you really think slop consumers are going to pay more to go through the trouble of migrating to a deGoogled phone? I paid more for a less usable phone OS because I hoped Murena would maintain some humanistic principles for longer than a year. Feel free to reach out if you find yourself in need of a good product and marketing consultant soon. I'm not cheap, but at least you can trust I won't bullshit you.

    💬 1🔄 4⭐ 23
  • 💬 0🔄 0⭐ 4
  • Nov 7, 2026, 9:37 AM

    @gael please don't be one of those tools who says "AI is just a tool". Yes, it is usefull in some cases, but in most cases it's not worth the cost. Just the environmental impact alone is bad enough.

    💬 1🔄 0⭐ 10
  • 💬 0🔄 0⭐ 0
  • 💬 2🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 8, 2026, 4:07 PM

    @gael @nini how? Are you going to train your own AI on data that isn't stolen? If so, will you do it on hardware that doesn't require vast resources and therefore doesn't emit any CO2? And how will you ensure your skills will not be deteriorated from relying on your "tool"?

    💬 0🔄 0⭐ 0
  • Nov 7, 2026, 12:16 PM

    @gael
    Define 'powerful'.

    From where I stand, it looks like at best LLM-based technology increases speed and reduces labour costs in a few business sectors at the cost of burning the planet and increasing every measure of inequality across the globe.

    I would expect an ethically driven company like @murena would have very few 'AI needs'.

    I hope you aren't following Andy Yen into #AIPsychosis and ruining a trusted platform.

    💬 4🔄 8⭐ 29
  • 💬 0🔄 1⭐ 8
  • larrylarrythecow@wetdry.world
    Nov 7, 2026, 12:28 PM

    @tomstoneham

    hm, it seems that among other things, the "elixir office" is completely vibe coded with the listed github contributors being 0% human

    @gael @murena

    💬 0🔄 0⭐ 1
  • 💬 2🔄 0⭐ 0
  • 💬 0🔄 0⭐ 1
  • Nov 8, 2026, 9:36 AM

    @gael

    Above quote concerns "elixir office" - "totally AI-coded under my direction".

    By doing this, you have damaged your brand. I have a FP5 for personal, and a FP6 for work. I would not have bought them if I had known the above. I accepted paying 3x the cost of a cheap android phone for a promise of 10 years support, and for a company with ethics.

    As it stands, my FP5 replacement, when necessary, will be from someone else.

    @larrythecow @tomstoneham @murena

    💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 6🔄 0⭐ 0
  • Nov 7, 2026, 12:56 PM

    @gael @murena I live on planet earth. The one where 24% of humans do not use the internet and 750 million don't have electricity.

    And in the UK where the goverment is allowing 6GW of new data centres (against citizen wishes) by 2030 to support AI ambitions even though peak national energy consumption in winter is 45GW. I.e. adding 13%. From where?

    It is also November and we still have water shortages in some parts of this famously rainy country.

    That is my world. That is why I #ResistAI

    💬 1🔄 10⭐ 22
  • 💬 1🔄 0⭐ 0
  • 💬 1🔄 2⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 0🔄 0⭐ 3
  • Nov 8, 2026, 8:11 AM

    @gael @tomstoneham @murena

    That is correct. In the real world, "AI" is nothing more than a toy that is fundamentally incapable of doing the things its proponents think it can. It's extraordinarily unethical in how it's created and the energy, water, and space requirements. It causes what amounts to brain damage in its users to the point that they quickly lose the ability to realize how utterly horrible it is.

    I guess this means that Murena and /e/OS are effectively dead now.

    💬 1🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • 💬 0🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • 💬 2🔄 1⭐ 1
  • Nov 8, 2026, 10:54 AM

    @gael @SuperMoosie @tomstoneham @murena no, it isn't. We should be desperately cutting energy consumption and fossil fuel emissions in order to survive, and you are touting a technology whose emissions already equal those of an average European country and are set to multiply. "Green" energy sources should be used to replace fossil energy for *essential* consumption purposes, not to serve a new technology that is an energy hog like no other. Completely irresponsible.

    💬 1🔄 1⭐ 3
  • Nov 8, 2026, 10:59 AM

    @titia @SuperMoosie @tomstoneham @murena ok but for this first stop to use a fuel car, stop flying planes, and convince thr US and China amd some other countries to stop producing electricity with fossils, bc they are responsible for the biggest carbon impact.
    Also I'd to inform you that Murena Workspace runs on an hydropowered DC, and that the rest is mostly powered by decarboned electricity (mix of solar, wind and nuclear)

    💬 1🔄 0⭐ 0
  • Nov 8, 2026, 12:51 PM

    @gael @titia @SuperMoosie @tomstoneham @murena Gaël.

    Not everyone thinks AI is a good thing.

    Even if they can't completely avoid it, because as you point out the Linux kernel now has AI slop, they want to at least avoid it as much as possible.

    And guess what?

    Many of the people who do are the same people who would consider buying a deGoogled phone.

    Anyone who wants AI features integrated into their phone would buy a Google Pixel!

    (1/2)

    💬 2🔄 3⭐ 9
  • Nov 8, 2026, 12:58 PM

    @gael @titia @SuperMoosie @tomstoneham @murena By arguing with your potential customers like this, you are risking seriously alienating a large segment of your potential customer base.

    I would very strongly recommend you stop, put your phone down, touch some grass, and have the humility to consider whether "new AI features" is really the message your potential customers really want to hear from you.

    (2/2)

    💬 0🔄 3⭐ 7
  • 💬 1🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • Nov 10, 2026, 1:51 AM

    @gael @usernomnomnom @titia @SuperMoosie @tomstoneham @murena We're talking about AI in /e/OS because the founder of /e/OS wrote: "AI is the most powerful technology ever available for humanity and I wish the good people embrace it too." (Among other pro-AI comments.)

    Which implies more AI features will be a part of /e/OS going forward.

    As you're probably now very aware, there's a significant proportion of the /e/OS and Murena userbase who are not exactly enthusiastic about AI. (1/2)

    💬 0🔄 0⭐ 0
  • Nov 8, 2026, 1:41 PM

    @ajsadauskas I use /e/OS on a Fairphone and a Volla Tablet. The idea that they're integrating AI into any part of the workflow fills me with terror because I made the choices I did SPECIFICALLY to support what I thought was an ethical set of companies who cared about the planet and the people on it. If Murena starts using AI, that absolutely leaves me in a hole with nowhere to go; it's not like I'm going to be able to keep investing money, time, and experience sharing with others (who I have been pushing to /e/OS as a safe and ethical alternative) - I refuse to push people to a place that embraces the planet-and-people-destroying technology that is LLMs and agentic AI.

    @gael @titia @SuperMoosie @tomstoneham @murena

    💬 0🔄 0⭐ 0
  • 💬 1🔄 0⭐ 0
  • Nov 8, 2026, 1:55 PM

    @gael @ajsadauskas @titia @SuperMoosie @tomstoneham @murena Even local LLMs and agentic AI are built on theft as they are based on large scraping models that also cause significant environmental harm for their continual training. Additionally, scientific research has shown that using LLMs and agentic AI causes a decline in the cognitive capability of the users - this is not something that anyone should be pushing for. We want to educate and uplift humanity and the planet, not destroy it.

    The terror comes at the thought that I may have unwittingly contributed to advancing the causes of pro-AI companies with not only my own resources (money, time, marketing) but also having encouraged others to use their resources as well. Because I believe in promoting the health of the environment and the overall improvement of humanity, the idea that work I've done has translated to the exact opposite effect is heartbreaking. That's what scares me the most.

    Edit: And I see above the "what about X," arguments that you're making. And you're not inherently wrong! We SHOULD be pushing to use less fossil fuels (ideally none), as well as a countless number of other things to improve the planet and our societies. But that's not a shield to say, "Well, other things suck more, so this thing that sucks is okay." It's ALL bad. We should push back against ALL of it, as much as we can.

    💬 0🔄 1⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 8, 2026, 6:27 PM

    @gael @tomstoneham @murena very common for capitalists to be tone deaf and live in an alternate reality, in the face of the harm they've created for the common person. If you consider cocaine a productivity metric because it lets you write code faster, then go ahead and embrace your snake oil.

    Nobody else around you sees anything but dead labour being used to insult the creativity and ingenuity of millions of people, watered down to a black box which can produce everything from technical debt to misinformation. I used to respect your projects, I was a Mandriva user for a decade. But in the end, tech bros use their privilege at the top of a project to let us know how "unenlightened" we are for not choosing the thing that is proactively being advertised as a way to destroy our livelihoods, our environment, our passions, for the sake of "being more productive." Sorry, not sorry, we'll gladly drawn the line in the sand here.

    💬 1🔄 1⭐ 0
  • 💬 0🔄 0⭐ 0
  • Nov 9, 2026, 10:49 AM

    @gael

    My dude, you aren't required to rebut every argument, or have an answer to any point.

    Either say you don't know, or say you don't agree. Or, say you don't care.

    Be honest instead of trying to protect your pride, moral reputation or more likely both.

    💬 0🔄 0⭐ 0
  • Nov 7, 2026, 7:41 PM

    @tomstoneham @gael @murena Imho agriculture is a much more powerful technology, it allows to feed billions
    Writing is another example...

    Super sad that e/OS and murena are also getting into that unethical shit

    💬 0🔄 0⭐ 1
  • 💬 0🔄 0⭐ 0
  • Nov 8, 2026, 11:16 AM

    @tomstoneham Had not heard of Murena.

    "Degoogled technology"

    Odd to read; the only reason I'm Degoogling is to get away from the LLMs.

    💬 0🔄 0⭐ 0
  • 💬 3🔄 1⭐ 14
  • 💬 8🔄 1⭐ 1
  • Nov 7, 2026, 1:12 PM

    @gael @djoerd oh, and I never use STT ;)

    Somewhat to my surprise, years ago I expected speech input to be of much more value than it turned out to be, really.

    I think the reason is that spoken input (and output) can be overheid by anyone near. I much rather type, even if on an annoying phone keyboard with errors in autocompletion 😬

    💬 0🔄 0⭐ 4
  • Nov 7, 2026, 3:01 PM

    @gael @djoerd calling everything AI is unfair. STT and LLM are order of magnitude different in terms of training AND inference power and lelory consumption.

    💬 0🔄 0⭐ 12
  • Nov 7, 2026, 3:53 PM

    @gael
    STT is fundamentally not the same thing as the original discussed subject, namely LLM powered IAGen. Using one to justify the other is just bad faith, especially for someone who understands that difference like you.
    @djoerd

    💬 0🔄 1⭐ 16