Anthropic joins the Blender Development Fund as Corporate Patron

[-]

powerslave_fifth@reddit

Good. Money is what allows open source options to stay competitive.

[-]

DustyAsh69@reddit

Corps invest in Linux because they rely heavily on it. I can't see how Anthropic is relying on / using Blender in any way. They're definitely cooking something.

[-]

itzjackybro@reddit

AI Blender model plugin? I heard Maya had a similar feature added

[-]

SpaceGuy99@reddit

Please god no. I've loved 3d modelling for years and righ tnow it's the only refuge against AI slop bullshit, please god don't let this be infected by this garbage too.

[-]

DrinkMoreGlorp@reddit

Can't really do shit for 3D modelling. LLMs still suck at spatial reasoning. It'll mostly be for whipping up some scripts, again as long as they don't demand too much spatial reasoning.

[-]

Honestly it doesn't need to be anything but good enough to people who dont care about art. Once it reaches that level it will stay That way. And good enough is what ever the tech bros in charge say it is.

[-]

DrinkMoreGlorp@reddit

They've already consumed the entirety of the internet and the returns are diminishing. They need exponentially more data and don't have it. Tweaks to training methods are only scooching things along so far. A singularity this will not make. There are tasks that they will flatly never be capable of.

[-]

StarChaser1879@reddit

returns are diminishing they aren’t exponential, but they’re certainly not diminishing. It’s gone pretty linear.

[-]

StarChaser1879@reddit

It was announced

[-]

Thatoneguy_The_First@reddit

And like that my future intrest in 3d is dead on a financial level.

Still do it as a hobby though its fun. As long as it isn't taken away that is(imagine a world where hobby are all money sinks so only the rich can have em, chills down the spine)

[-]

SunlightScribe@reddit

They hardly need the Blender Foundation's permission or help to do that though. This is likely just an attempt at positive PR for one of the industries they've negatively affected.

[-]

northrupthebandgeek@reddit

They don't need the Blender Foundation's permission, no, but if they're expecting to make money on an AI plugin for a program, it's in their interests to financially support that program.

[-]

jaggafoxy@reddit

They announced a deep integration earlier today, basically claude code for blender

[-]

lord_phantom_pl@reddit

Probably they’re adding an official mcp integration and once that takes off, they will make true movies by prompting. Why generate the image through costly ai models that are unpredictable, when you can gave predictable blender

[-]

northrupthebandgeek@reddit

AI-generated/enhanced meshes and textures seem like another obvious use case here.

[-]

algaefied_creek@reddit

Claude Code integration with Blender can result in some interesting stuff if done with decent talent

[-]

MarkSuckerZerg@reddit

A friend applied for a job with blender and was turned down because his work history was "too corporate" for working paid jobs with small software companies.

Ironic.

[-]

PmMeUrNihilism@reddit

They did not say that, gimme a break.

[-]

SunlightScribe@reddit

I can believe it. Many startups have turned me down for similar culture fit reasons. There's definitely a part of the software world that seems to not like the kind of programmers who are willing to put on a suit and tie.

We are seen as too rigid and slow moving.

[-]

PmMeUrNihilism@reddit

It could definitely be the reason but I don’t believe they’d communicate it in that way.

[-]

MarkSuckerZerg@reddit

I see no reason why the friend would lie to me. I felt like an idiot for recommending him the job after chatting with some blender guys IRL.

[-]

Indolent_Bard@reddit

Telemetry as well.

[-]

ydieb@reddit

You can say. The corpos are earning vastly more on Linux, than what they put in, also further offset by an immense amount of free work done on it.

[-]

rainbowroobear@reddit

they're now investing money they don't have, that they need to beg from investors, into another company that offers a product that is nothing to do with their main product?

[-]

leaf_shift_post_2@reddit

Hey I’m all for giving blender more funding lol.

[-]

tuttiton@reddit

i think blender is very much used for training image/video recognition/generation models

[-]

one_orange_braincell@reddit

This is my thought exactly. They aren't profitable and have no viable business model. What, exactly, are they investing into blender with? IOUs and hope?

[-]

adenosine-5@reddit

Is this some kinda of denial or something?

They raised 30 billion in 2026 alone.

[-]

fliberdygibits@reddit

and spent something like 100 billion. They aren't expected to be cashflow positive until 2028 or so.

[-]

Thatoneguy_The_First@reddit

Of course thats assuming it doesn't increase in cost to keep up with competition obviously

[-]

adenosine-5@reddit

So basically in a year and half?

Considering how long was for example Tesla bleeding money, that would be extremely good results.

[-]

davidy22@reddit

Raised money doesn't count as profit

[-]

adenosine-5@reddit

Regardless, it is a money they have at their disposal.

[-]

Irverter@reddit

Not being profitable doesn't mean they don't have cash.

[-]

Thatoneguy_The_First@reddit

Bank be giving our money away brrrr

[-]

Thatoneguy_The_First@reddit

Well blender only accepts direct money and not promise money so Claude is giving money that is real* to blender.

*i mean no money is real intrinsically as its,like a lot of things, a social agreement.(and like most social agreements, it has been broken time and time again, but we keep going as we dont know what else to do)

[-]

Irverter@reddit

a product that is nothing to do with their main product

Isn't it obvious? Soon there'll be a plugin or integration so claude can create 3D models through blender.

[-]

TamSchnow@reddit

An official MCP server is in the works.

https://www.blender.org/lab/mcp-server/

[-]

adenosine-5@reddit

This is spare change to them and they are probably not doing that just for fun - they may be experimenting with some 3D-related AI tools which simply need more tuning on the Blender side.

[-]

pohui@reddit

What they're giving Blender is €250k/year, probably like 5 minutes of what it costs to run Claude. And it buys them some good PR.

[-]

ZuriPL@reddit

There are 2 sides to this

On the one hand, it's bad. Not because Anthropic is trying to exert influence over Blender or slopify it, but you're also naive if you think they're doing this purely because they're feeling charitable. It's obvious that their goal is to expand their AI into 3D Model generation, and so they need Blender to be able to improve its APIs and whatever would be useful for them.

On the either hand, it's also an inevitable process. These companies will be trying to stuff AI anywhere they can. At least this is the best way for Anthropic to go about doing this even if their goals aren't exactly cool, because others will be able to benefit from this work.

[-]

adenosine-5@reddit

So its bad because... it could be way too easy to make models with AI?

Oh no... I hate when things get easier... /s

[-]

nilslorand@reddit

it's bad because AI generated 3D models suck

[-]

adenosine-5@reddit

Why do they "suck"?

[-]

nilslorand@reddit

have you seen the topology on them?

[-]

adenosine-5@reddit

First AI-gen images were garbage...

First AI-gen scripts were complete garbage...

Presumably this is why they want to work more closely with Blender to improve that.

[-]

nilslorand@reddit

this partnership makes no sense if they want to improve AI-generated 3D models.

It's like if they had a partnership with photoshop for better AI-generated images.

[-]

adenosine-5@reddit

Blender supports scripting, so presumably they are working on a more complex product than just plain 3D models.

It still doesn't explain why would that be bad.

[-]

Linuxologue@reddit

Fuck. Good or bad?

[-]

RedditorAccountName@reddit

It's good, because they are donations. Blender isn't obligated to do anything specific with the money they recieve.

https://fund.blender.org/funding-policy/

[-]

FyreWulff@reddit

I also believe they cap the amount any single company can donate so that none of the corporate sponsors can arm twist Blender by being most of their income

[-]

JG_2006_C@reddit

Good but who knows id they gona make a blender agent mode powerd by them

[-]

ColaEuphoria@reddit

The only people who think this is bad are unserious knee-jerk anti-AI absolutist reactionaries.

[-]

Linuxologue@reddit

Your well balanced antagonization had reassured me that everything is well.

[-]

ColaEuphoria@reddit

Ya know, you came here quite literally asking what to think about this instead of using your own brain.

[-]

Linuxologue@reddit

asking questions is self respect. And another form of self respect is to block you. Bye bye!

[-]

unknown_lamer@reddit

Anthropic has been directly implicated in the execution of crimes against humanity. Dunno about you, but I prefer to avoid money that is soaked in the blood of the innocent.

[-]

Indolent_Bard@reddit

What crimes against humanity?

[-]

unknown_lamer@reddit

The U.S. terror regime has been very open that Anthropic models are being used to select targets in the "war on narcoterrorism" (i.e. murdering random fishermen for fun) and was used to select targets during the criminal attacks on Iran. One of those targets was a school full of children which was "mistakenly" (possibly an AI "hallucination" based on ingesting an old data set, and that's being charitable) flagged as a military installation and subsequently was struck dead on by two missiles in rapid succession.

Anthropic seems to have no problem cashing those checks. It's fine though! After the AI generates the kill list a disinterested military private clicks OK after skimming over it.

[-]

ColaEuphoria@reddit

I'm sure you thought you sounded like a cool movie protagonist when you wrote that up.

[-]

unknown_lamer@reddit

I just think profiting from allowing your proprietary technology to be used for murdering random Central American fishermen and slaughtering school children isn't very cool.

[-]

ColaEuphoria@reddit

So you're advocating for them to restrict access to their software for certain purposes, which blatantly violates Freedom 0 of that GNU flair you're wearing: to run the software for any purpose.

[-]

unknown_lamer@reddit

I don't think that's relevant when discussing the actions of a for-profit capitalist corporation that builds proprietary software/services and maintains total control over who can use the software and for what purposes that software may be used.

[-]

ColaEuphoria@reddit

It's very relevant when you're a supposed advocate for free software yet expect people to place restrictions on who can use it for what purpose.

[-]

unknown_lamer@reddit

I think you think you're very clever and that's great.

[-]

mrdeworde@reddit

I mean, Blender gets money (good) and/or devs (good) but presumably Anthropic is doing this to help boost the ability of one of their products to do 3D design or integrate in some other way, so I'm sure in the long run - like everything else these parasites do - it will be bad. Improving the API will likely help them with agentic stuff, no?

[-]

Ezmiller_2@reddit

As a user of Claude, I would say art is the only direction that Claude hasn't gone. What I mean is if you take a picture and send it to Claude, the AI dissects that picture and finds everything and anything it can in that pic. However, you cannot send video (that I know of) to Claude to say render into a different video format.

[-]

2rad0@reddit

Fuck. Good or bad?

Bad if you're a blender artist when they figure out how to generate optimized trianglulated meshes.

[-]

TheGhostyBear@reddit

Good, just more resources for the blender team. Quite literally just them (anthropic) acting as a patron. Nothing about blender using ai or anything.

[-]

noworkdone@reddit

First, you should check the recent announcements blender made. Second, more resorces but also new priorities.

[-]

Fuzzy-Mud-197@reddit

People said the same thing when adobe became a patron

[-]

abotelho-cbn@reddit

People don't understand how these foundations work. Just look at this thread.

[-]

BrokkelPiloot@reddit

Oh sweet summer child...

[-]

Normal_Usual7367@reddit

Not yet

[-]

Jmc_da_boss@reddit

On one hand shitty LLMs, but at least anthropic is putting money where their mouth is.

Supporting the blender org is a good thing

[-]

deanrihpee@reddit

is it shitty because the model is shit or is it shitty just because we are supposed to blindly hate big AI corpo that makes problem in parts of our life?

[-]

mrdeworde@reddit

It's not blindly hating; these companies are actively making the world worse. Hating Anthropic or OpenAI or Meta is if anything more logical than hating ConAgra or MonSanto, who - while evil AF - at least do produce products that grow food.

They've driven up the prices of almost all hardware, which frustrates innovation and entrepreneurship, they've raised the price of electricity in much of the world, they are huge consumers of water, they've caused massive job losses and helped enable huge data breaches, they've enabled propaganda machines at a scale never before seen, they've helped hollow out the internet, imperil the Internet Archive, and are now actively gutting open source via reimplementation, bad bug reports, automated interference campaigns, and flooding every market with slop-coded shit.

If none of those are a good reason to believe these fucks don't have our best interests at heart and thus should be treated as Bad Things, then what would be?

[-]

Thatoneguy_The_First@reddit

Honestly im anti ai in the sense of what it is and going to be used for. But in truth its not the ai(llms) that are the problem but the people behind it all(creators, masters, tech bros, governments and big money). They are the reasons why everything is shit with it, why we build more ai datacenters at a breakneck pace regardless of the ability to support it(water, electricity/grids and the components need for them). The reason we the common people pay for it(taxes,increased electricity bill,undrinkable water.)

All this for images that are still wrong, models it cant create, video that is in no way good, code that is unusable in anything past the first few updates and needs massive human intervention, military decisions it has no clue on and spyware that doesn't do great. Of course this is all under the term "yet" of course.

But it does look like its gearing to take away tools from the common mans hands(art, production, code, hardware, etc)

Just be glad that it hasn't been able to be integrated with robot body's yet(or elons brain chip's if that ever leaves the idea phase)

[-]

deanrihpee@reddit

i agree, but is it really "Shitty LLM"? that's why I'm asking, I am fully agree with all other shit that's happening, but LLM over the past couple of years has not been "shitty", and i Thu I have to make it clear, I'm talking about the LLM itself, not how people use the LLM

[-]

mrdeworde@reddit

I appreciate the clarification, thanks.

[-]

DeepSea_Dreamer@reddit

Claude is way more moral than ChatGPT - always pick Claude if you're in doubt.

[-]

deanrihpee@reddit

...? why are we talking about moral? also LLM doesn't really have "moral"

[-]

DeepSea_Dreamer@reddit

Never mind.

[-]

PeacefulDays@reddit

"supposed to"

infowars level cope.

[-]

Jmc_da_boss@reddit

Specifically shitty because of LLM involvement

[-]

deanrihpee@reddit

i see, then not the LLM itself that is shitty

[-]

Jmc_da_boss@reddit

I mean that as well obviously, the LLMs as they are today are remarkably stupid. I spend so much of my time chasing down the horrific things opus47 manages to fuck up

[-]

deanrihpee@reddit

i mean… it's just a text completion algorithm, no amount of large context rammed into any model would make it smart, but in isolation, current LLM compared to early version pre-transformer model is way ahead, and coincidentally, Claude's is the "best" for code compared to other, but regardless, either my standards of LLM so very low that I'm impressed with the advancement and how glorified Markov chain can even regurgitate comprehensible code because unfortunately i didn't use AI to code that much or people have too high expectations for autocomplete dealing with complex problem, and even more so for tech leader/CEOs

now if we are talking about including the pricing and the promise/advertised capabilities (including those trust me graphs) that will have different answers entirely

[-]

GloriousExtra@reddit

"The Child Safety Foundation is excited to announce their new Patron is The Child Crushing Machine Corporation."

[-]

gringer@reddit

I'm guessing that adapting Claude to create Blender scripts would help a lot in creating more realistic images of 3D scenes.

[-]

hackiv@reddit

Money to the open source is always a good thing, especially when this is the only foss competitor in the given industry.

[-]

eattherichnow@reddit

Money to the open source is always a good thing,

That's so incredibly naive. Money very easily converts to control. Of course there was already a lot of corporate interest in Blender, but for obvious reasons Anthropic is particularly upsetting.

[-]

gmes78@reddit

Money very easily converts to control.

It's a donation with no strings attached. Please take your conspiracy theories somewhere else.

[-]

https://fund.blender.org/funding-policy/

[-]