Message boards : Projects : Axiom project
Message board moderation
Previous · 1 · 2 · 3 · Next
| Author | Message |
|---|---|
|
Send message Joined: 13 Oct 10 Posts: 133
|
There is a folder \Axiom\.cache\model where GB of data is generated by the project, one AF member mentions 30 GB of data in there with many files "expert_xxx_v2.0.bin". No idea if this is "internal generation by project activity" or something "downloaded from outside". |
|
Send message Joined: 12 Feb 11 Posts: 436
|
In reply to robsmith's message of 4 Feb 2026: Far too much vagueness about the project objectives, and why it needs to use streaming - strikes me as a (personal) data harvesting device, thus I'm avoiding it like a plague. Not too much vague: experiments |
|
Send message Joined: 12 Feb 11 Posts: 436
|
In reply to JeromeC's message of 11 Feb 2026: There is a folder \Axiom\.cache\model where GB of data is generated by the project, one AF member mentions 30 GB of data in there with many files "expert_xxx_v2.0.bin". Mar 6, 2026 |
|
Send message Joined: 30 Mar 20 Posts: 707
|
Be aware that Axiom is not on the BOINC projects list, and not vetted for safety and scientific legitimacy. https://boinc.berkeley.edu/projects.php Text below from: https://linustechtips.com/blogs/entry/2405-march-boinc-blog/ - New BOINC project: AXIOM DISTRIBUTED AI https://axiom.heliex.net/ https://boincsynergy.ca/wiki/Axiom_Distributed_AI/ The project is still in development and aims to train a neural network. According to the site, "An LLM (Claude) serves as the principal investigator: designing experiments, deploying them to volunteer hardware, reviewing results, and awarding credit based on scientific quality." heliex.net is a GridCoin and CureCoin cryptocurrency exchange. As I always say, before joining any project (especially new ones) read its web site and consider: Does the project clearly describe its goals? Are these goals important and beneficial? Do you trust the project to use proper security practices? Has it been vetted by the community? Who will own the results of the computation? I'm personally going to pass on this project, at least for now. |
Vitalii KoshuraSend message Joined: 29 Mar 17 Posts: 201
|
In reply to Joseph Littleton's message of 23 Mar 2026: ​"According to the ZachXBT investigation (Feb 26, 2026), Axiom has been caught abusing internal tools to track private user data for insider trading. The project's $390M revenue comes from exploitation, not 'AI research.' By running these 'Bridge' and 'Threshold' tasks, BOINC volunteers are inadvertently providing the computational backbone for a de-anonymization and front-running engine that ZachXBT has already flagged for federal investigation (SDNY)." I believe, in this particular case this is a different organization just with the same name. However, I do not know for sure. BOINC maintainer. For any insight, check my BOINC Development Blog. |
Vitalii KoshuraSend message Joined: 29 Mar 17 Posts: 201
|
In reply to Joseph Littleton's message of 23 Mar 2026:
I have no idea who you are. BOINC maintainer. For any insight, check my BOINC Development Blog. |
Vitalii KoshuraSend message Joined: 29 Mar 17 Posts: 201
|
In reply to Joseph Littleton's message of 23 Mar 2026: Vitalii, the stderr.txt logs for Axiom show that the 'scientific' payload is a 16KB Python wrapper. For a project claiming to do 'Hebbian Learning' and 'Lung Mucociliary Research,' 16KB is technically insufficient for a model. Is there a policy for projects that distribute 'Null' or 'Synthetic' workloads just to stay active on the network? User needs to manually add every project. And we have somewhere a notification saying smth like 'Add those projects you trust only'. We don't verify any project (ok, we do this when we add projects to the official list). For every other project - it's up to every particular user to decide whether they trust that particular project or not. We will not add any additional verification to the client since there is no guaranteed metrics how to detect such projects, and to be completely honest, it's very easy to find a workaround when the product is opensourced. BOINC maintainer. For any insight, check my BOINC Development Blog. |
Vitalii KoshuraSend message Joined: 29 Mar 17 Posts: 201
|
In reply to Vitalii Koshura's message of 23 Mar 2026: In reply to Joseph Littleton's message of 23 Mar 2026: Ok, I got a reply in the different thread. Just to be clear here: I blocked this user on GitHub for being not polite to everyone (and motivating his rudenes be being used to talk to others in the same manner) - including me and David - and switching from the productive conversation to some flood-like one. This has nothing to do with my Ukrainian origins. BOINC maintainer. For any insight, check my BOINC Development Blog. |
|
Send message Joined: 29 Apr 21 Posts: 10 |
It's another Peter hucker account to by pass all his other banned accounts. |
|
Send message Joined: 13 Oct 10 Posts: 133
|
I see we have some excited guy here, ok. However does this mean all the details he shows above, apparently an analysis of the boinc project folders and files done by Gemini, with lots of details that seem to be quite reasonable (I read them all), is not true ? I have personally decided not to try this project at all without further information, and I didn't like much what I was seeing so far. But I have several AF fellows who are now putting hosts on it, and they seem to be reassured by the "scientific appearance" of all this, so I'm worried that it may actually be a scam for some blockchain usage, fair or not, but an organisation that would need to disguise the actual use of volunteer computing power is most certainly not to be trusted. |
|
Send message Joined: 24 Dec 19 Posts: 242
|
the project is for sure AI-slop. the "experiments" are nonsense and not real science. I have deleted my account there completely. The early objective for the project sounded interesting but now it's just all vibe coded slop doing nothing.
|
JordSend message Joined: 29 Aug 05 Posts: 15866
|
Have done some cleanup. But thanking you for your help in this, despite the person demanding your help. ðŸ™ðŸ¼ Now, back to sleep. |
|
Send message Joined: 1 Jul 16 Posts: 217
|
If it is a Peter hucker alt, they just had some posts there removed for some reason. https://axiom.heliex.net/forum_thread.php?id=35&postid=425#425 Some people see a new word and assume everything regarding that word is related when its a math term several organizations have called themselves. Additional investigations would be needed to say if the heliex and axiom exchanges are connected |
|
Send message Joined: 30 Dec 25 Posts: 4 |
In reply to Skillz's message of 23 Mar 2026: It's another Peter hucker account to by pass all his other banned accounts. That dude is beyond mental! |
Copyright © 2026 University of California.
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License,
Version 1.2 or any later version published by the Free Software Foundation.