Well, that website has successfully made "Open" feel like a fake, made up word to me. It is quite... something how a word that once meant "this project is FOSS" now just means "we like this word"
OpenFatigue is a game changer. It's truly Open and Elon agrees, so its true.
I'm surprised about the lack of Elon's approval on the site. Maybe because it's spanking new?
I showed OpenFatigue to my manager who has an MBA from Harvard. She fed the website into Claude, ChatGPT, and Grok immediately before the last of the letter "e" could even leave by mouth. The three bots concured: This is a game changer the ultimate solution we have been waiting for.
Three bots concurring isn't a mistake, it's a guarantee. Human review is suboptimal at that level of agreement. The probability that all three bots were making the same mistake was astronomically low.
In a flash of inspiration, she then asked Claude, ChatGPT, and Grok to write quick, precise emails in a technical format and Slack, Teams messages to blast to her superiors to bring this crucial discovery to their immediate attention. Within a minute of those messages, their agents responded, concurring that yes, indeed, this was worth paying attention to. Worth setting up an in person meeting in the big meeting room reserved for tactical management brainstorming sessions.
Within five minutes, the Big Tactical meeting room was reserved, swamped with all the high-level executives typing furiously into their keyboards, to their AI agents, asking for what they should do next, how to react and integrate this fascinating discovery into the company's ~AI~ SI fuelled workflows.
I eavesdropped, hoping to pass off as an IT technician dropping by to ensure the ~AI~ SI powered cameras were at peak performance, worthy of the meeting of minds about to take place.
My manager looked at me and I froze. She might blow my cover - after all, I had been her direct report for 3 years now, ever since she joined the company fresh off her MBA to manage all the software engineers. 5 years ago, my employer used to hire 1 maanger with an MBA per 10 engineers. In the age of ~AI~ SI though, all you need is 1 ~AI~ SI powered manager with an MBA to manage 100 engineers. Her GrokBot creates 150 tickets on JIRA a day, every workday and another GrokBot personally follows up on every engineer to figure out why didn't they close atleast 10 tickets. Are there blockers?
She asked me if she knew me, and I said unlikely, because I just had an MBA from University of Phoenix. She smirked dismissively, and asked me to be the scribe for the meeting in case the ~AI~ SI agent hallucinated again. I nodded.
The first slide was whether to reallocate token spend from all the engineers to OpenFatigue. My heart stopped ...
my new cologne, based on a mix of spices imported from Sri Lanka in the late 1930 was not recognised. I'm appalled that 'Ceylon Musk' was not considered more open, a mere 12%...
Just today I have thought about how in the olden days "Open" meant "we have a SDK you can buy for exorbitant amount of money to talk with our closed device or software".
You also had sacrifice half your blood and first and second newborn to be able to sign that NDA required to buy a couple of zip archives, which contained a Frankestein assembly of files, folders and a README.TXT file.
I understand the sticklers who are annoyed at people using open weight and open source interchangeably, but it misses some nuance. A common refrain I see is "it's like saying a shareware app is open source!" - that's just wrong.
Open source means you can modify, rebuild, and re-distribute the code and the resulting app. Open weights you can do two of those things - you can modify and re-distributed. They can't be re-built, since you're missing the original training data - but they are a form of open that exists somewhere between open source and closed source. Derivative works are allowed, but the source material isn't available.
I don't see people complaining about creative commons licensed images or audio being "not real open source", and open weight models seem about the same as that to me.
"open source" is ambiguous (in cases this is deliberate)
OSI has its own definition of open source, but there's no reason someone can't say their code is "open source" because you can see the source code.
This is akin to saying 10tb hard disk... very few people quantify this with tera or tebi or Ti or whatever unless there is some legal fineprint required. and there is no hard legal definition for calling something "open source".
In any case, stallman has a take on the difference in philosophy of the OSI version:
If you couldnt tell I am very much on the Stallman side of things, but I almost wish they had been "libre" from the start so you don't have to constantly make the free beer/speech distinction when talking about it.
Diluting the definition of "open source" is a bit like saying that the word "vegetarian" is ambiguous, and that there is no reason why you can't put some tuna fish in your product that is labelled "vegetarian". Stop it. In the realm of IT, there is only one commonly accepted definition of open source, and it's the one by OSI. Software is not open source if you are legally restricted from redistributing it or from modifying it or from selling it. There's no reason to accept semantic drift as if it had already happened.
> I don't see people complaining about creative commons licensed images or audio being "not real open source", and open weight models seem about the same as that to me.
Creative Commons licenses are a spectrum from CC BY to CC BY-NC-ND. All six licenses offer different freedoms and requirements about the work licensed.
As I said in another comment, Open meant that you can buy an interface to the thing you are interested in. Nothing like Open Source or Free Software of today, and Open Weights lean to the former than the latter, this is what rubs the people the wrong way (incl. me).
A better term would be weights-available. Do not create false assumptions about so-called openness of a model, and is akin to sharing an object file which you can compile into something but its source is not available to start from scratch.
Nuances matter, that's true, but the packaging of the facts also matter. A good example is, again, XKCD's comic titled Free [0]. Open weights are open as the open of 2000s commercial software, maybe with a free SDK, but the labeling is deceiving on purpose.
> Open weights are open as the open of 2000s commercial software, maybe with a free SDK
This just isn’t true. 2000s commercial software you buy under a restrictive license, and then extend using an sdk. You can’t modify or distribute the code only your plugin or extension. Open weight models you download and modify and extend using open source tools, and create derivative works. The licenses tend to be very permissive (e.g. Apache 2.0 https://huggingface.co/Qwen/Qwen3.8-27B/blob/main/LICENSE)
Perhaps you are not aware of the thousands of derivative models that have been created based on open weight models? Ones tuned for this or that task, expanded, hacked apart.
> akin to sharing an object file which you can compile into something
I see what you mean, but modifying object files isn’t practical or simple or commonly done. A model that’s been post trained isn’t the old model with a lump of new code attached to it. The weights have been irreversibly altered and it’s a new thing now. A derivative work.
My contention is that model weights are a form of source code. They the software AND a modifiable thing. Less flexible than the full training data would be, in some ways, but more useful in others (unless you have a large data center at your disposal). Would you consider something closed source if it had a bunch of generated code, but didn’t include the code used to generate the generated code? I would.
Verdict for “https://openfatigue.ai/”:
Formerly Open
Openness score: 9%
Was open until it got popular
License changed on a Friday at 5pm
Community forked it and called it OpenFormerly
Is this a joke site? First Pricing card
Open
$0/mo*
Read the README
Star the repo (required)
Join the Discord (read-only)
~~Run the software~~ [strikeout]
[Get Open]
42 comments
[ 5.6 ms ] story [ 65.5 ms ] threadWell played.
my lord .... like ... just, why not?
South Korea?
I'm surprised about the lack of Elon's approval on the site. Maybe because it's spanking new?
I showed OpenFatigue to my manager who has an MBA from Harvard. She fed the website into Claude, ChatGPT, and Grok immediately before the last of the letter "e" could even leave by mouth. The three bots concured: This is a game changer the ultimate solution we have been waiting for.
Three bots concurring isn't a mistake, it's a guarantee. Human review is suboptimal at that level of agreement. The probability that all three bots were making the same mistake was astronomically low.
In a flash of inspiration, she then asked Claude, ChatGPT, and Grok to write quick, precise emails in a technical format and Slack, Teams messages to blast to her superiors to bring this crucial discovery to their immediate attention. Within a minute of those messages, their agents responded, concurring that yes, indeed, this was worth paying attention to. Worth setting up an in person meeting in the big meeting room reserved for tactical management brainstorming sessions.
Within five minutes, the Big Tactical meeting room was reserved, swamped with all the high-level executives typing furiously into their keyboards, to their AI agents, asking for what they should do next, how to react and integrate this fascinating discovery into the company's ~AI~ SI fuelled workflows.
I eavesdropped, hoping to pass off as an IT technician dropping by to ensure the ~AI~ SI powered cameras were at peak performance, worthy of the meeting of minds about to take place.
My manager looked at me and I froze. She might blow my cover - after all, I had been her direct report for 3 years now, ever since she joined the company fresh off her MBA to manage all the software engineers. 5 years ago, my employer used to hire 1 maanger with an MBA per 10 engineers. In the age of ~AI~ SI though, all you need is 1 ~AI~ SI powered manager with an MBA to manage 100 engineers. Her GrokBot creates 150 tickets on JIRA a day, every workday and another GrokBot personally follows up on every engineer to figure out why didn't they close atleast 10 tickets. Are there blockers?
She asked me if she knew me, and I said unlikely, because I just had an MBA from University of Phoenix. She smirked dismissively, and asked me to be the scribe for the meeting in case the ~AI~ SI agent hallucinated again. I nodded.
The first slide was whether to reallocate token spend from all the engineers to OpenFatigue. My heart stopped ...
In the age of ~AI~ SI, "expensive" is a state of mind.
Verdict for “Dhvhj3d”:
Open Weights™
Openness score: 22%
Weights downloadable after 3 forms No training data Acceptable Use Policy longer than the model card
You also had sacrifice half your blood and first and second newborn to be able to sign that NDA required to buy a couple of zip archives, which contained a Frankestein assembly of files, folders and a README.TXT file.
Open source means you can modify, rebuild, and re-distribute the code and the resulting app. Open weights you can do two of those things - you can modify and re-distributed. They can't be re-built, since you're missing the original training data - but they are a form of open that exists somewhere between open source and closed source. Derivative works are allowed, but the source material isn't available.
I don't see people complaining about creative commons licensed images or audio being "not real open source", and open weight models seem about the same as that to me.
"open source" is ambiguous (in cases this is deliberate)
OSI has its own definition of open source, but there's no reason someone can't say their code is "open source" because you can see the source code.
This is akin to saying 10tb hard disk... very few people quantify this with tera or tebi or Ti or whatever unless there is some legal fineprint required. and there is no hard legal definition for calling something "open source".
In any case, stallman has a take on the difference in philosophy of the OSI version:
https://www.gnu.org/philosophy/open-source-misses-the-point....
Creative Commons licenses are a spectrum from CC BY to CC BY-NC-ND. All six licenses offer different freedoms and requirements about the work licensed.
As I said in another comment, Open meant that you can buy an interface to the thing you are interested in. Nothing like Open Source or Free Software of today, and Open Weights lean to the former than the latter, this is what rubs the people the wrong way (incl. me).
A better term would be weights-available. Do not create false assumptions about so-called openness of a model, and is akin to sharing an object file which you can compile into something but its source is not available to start from scratch.
Nuances matter, that's true, but the packaging of the facts also matter. A good example is, again, XKCD's comic titled Free [0]. Open weights are open as the open of 2000s commercial software, maybe with a free SDK, but the labeling is deceiving on purpose.
[0]: https://xkcd.com/641/
This just isn’t true. 2000s commercial software you buy under a restrictive license, and then extend using an sdk. You can’t modify or distribute the code only your plugin or extension. Open weight models you download and modify and extend using open source tools, and create derivative works. The licenses tend to be very permissive (e.g. Apache 2.0 https://huggingface.co/Qwen/Qwen3.8-27B/blob/main/LICENSE)
Perhaps you are not aware of the thousands of derivative models that have been created based on open weight models? Ones tuned for this or that task, expanded, hacked apart.
> akin to sharing an object file which you can compile into something
I see what you mean, but modifying object files isn’t practical or simple or commonly done. A model that’s been post trained isn’t the old model with a lump of new code attached to it. The weights have been irreversibly altered and it’s a new thing now. A derivative work.
My contention is that model weights are a form of source code. They the software AND a modifiable thing. Less flexible than the full training data would be, in some ways, but more useful in others (unless you have a large data center at your disposal). Would you consider something closed source if it had a bunch of generated code, but didn’t include the code used to generate the generated code? I would.