Hacker News Reader

Comment by mittensc | Hacker News Reader

> Mozilla is focused on deploying its roughly $1.4 billion worth of reserves to support “mission driven” organizations, according to a new report.

> The nonprofit, also the parent of Firefox, is investing in artificial intelligence startups that are working on safety and governance issues in AI.

Why?, they want to go bankrupt?, do they like burning money?

I would understand investing in AI Tech... Brilliant if they use Mozilla contributors.

I would understand but investing in other startups... with due diligence and something that might make a difference

> that are working on safety and governance issues in AI.

what... why... what the hell... that's governments job, not mozillas...

View on HN

> that's governments job

Why do people take these AI "safety" research projects at face value? The real reason you need AI that is "safe" and "governable" is so that when you start having it promote advertisers content or support the current administration, you don't have to worry about it going "off the rails" and promoting a competing product to criticizing the administration.

I'm sure plenty of researchers in this space also believe they are working for the good of humanity, but I suspect the real am is much more practical and perfectly aligned with the business interests of all the companies sponsoring this type of work.

Uh, I mean, I think we've seen first hand that generally one aspect of safety we want is to not produce sexualised images of children.

Is it even an attainable goal? It seems an NN with less than say 4 billion parameters will be able to do that. The cost of training will likely go down with more models being available. Unless we lock down the computing for majority of people, I don't see how we can prevent someone creating a CSAM model in their garage.

I don't want to see CSAM created, but the totalitarian control required is too much for my taste (and frankly it's preferable for that person to use NN than to go out and hurt actual children).

Not to mention even locked down technology is often being abused by the privileged.

People can draw naked children with a stick and some dirt. I’m not sure that preventing the creation of fictional csam is the best use of our resources if we want to protect minors from abuse.

The best use? Probably not. But if I built a website that let people generate extremely convincing unlimited photos of you wearing an SS uniform and forcing your dog to smoke meth and sent them to everyone you’ve ever met, this might seem like a less worthy hill to die on. Or is that just a sticks and dirt thing too?

Oddly specific... Did that really actually happen to you?

Everybody I care about would know that those pictures are not real so I think that the harm to me would likely be lower than the harm to society if building websites were impossible.

The people who are sending the pictures are criminally liable, regardless where they got them.The fact that somebody built a website for it is irrelevant, the act of sending them unsolicited is the immoral act here. (And frankly it's probably gonna be laughed off or end up as spam unless somebody you associate with is an idiot.)

The goal is not to prevent someone from making their own model do what they want, but to prevent your model from doing what you don't want, like generate CSAM or non-consentual sexually explicit photos.

I don't understand how its controversial that someone/some-company might not want their products to be known for that.

It's a bit odd requirement, but.. OK. I mean who is the malicious actor here, the AI, the human user or the AI provider?

If the AI, then we shouldn't give it an agency (user should always vet the output). If it's the user, the AI is irrelevant to the question. And if it's AI provider, why would they train AI on such materials in the first place.

The whole enterprise of this kind of safety doesn't make much sense to me. If the AI is not able to follow so clear user instructions it's not ready for prime time and must be under human supervision at all times. (And subtle hallucinations on topic seem to be both more dangerous and more of a problem than blatant random production of explicit images, anyway.)

> It's a bit odd requirement

It is a realistic requirement for dealing with user generated content at scale, because this is a realistic requirement thing some people will do.

We’ve seen this fold out over the past month from Grok.

We currently can't do "AI safety" even in bleeding edge alignment research so investing into some startups in that area is just burning money. Current LLMs/ViTs have non-zero probability of producing something unsafe and it's their inherent trait.

> Current LLMs/ViTs have non-zero probability of producing something unsafe

Show me on the LLM where the "unsafe" result is that doesn't already exist inside its training data set? Go ahead I will wait.

Just so I understand your comment, are you trying to say that the money on safety should go first into research? And not into a later startup, which is just putting lipstick on the pig? I agree to this understanding, no doubt about it.

Basically yes. First figure out how to do reliable models that can be well-aligned with safety expectations, then invest into a startup that brings that tech to the industry.

> promote advertisers content ... promoting a competing product

Where is the value of AI when the responses are compromised like this? I could say the same thing about Google Search, which is one of the reasons I stopped using it.

Are we betting on the masses not caring that they're being lied to for profit?

They want NGO grant money. They look at the latest and greatest buzzwords for government policy spending and tailor their efforts towards acquiring that money, 99% of which will go to salaries and bonuses, 1% of which will be spent on the mission du jour.

Mozilla is a deeply corrupt and failed organization.

> Mozilla is a deeply corrupt and failed organization.

Can you elaborate?

See: Firefox, management churn, alienation and discarding of community, etc.

Some of us who donated money and supported Mozilla and Firefox are deeply, deeply disappointed and disgusted.

Principles are meaningless to non-human corporate entities, and I'll never donate to a non-profit, charity, or other institution again for the rest of my life.

A great deal of greed exists within many non-profits. (It's frankly obscene when you do your research.) That's not to say some don't serve the public well, but the legal structure of a non-profit isn't by itself enough to deter corruption.

I'm sure a lot of shady activity can be found if someone does some kind of bulk analysis on IRS 990 forms.

They're already an NGO who gives out grant money. When you say 99% will go to salaries and bonuses, you know all the financials are right there on the page right?

But compromised Mozilla's solution will then be passed as 'independent' so that a corrupt government can accept it without officially kneeling to BigTech. Publicity stunt a'la foundation.

That kind of reserves, invested wisely, would net enough interest to pay for a decently sized team maintaining Firefox and small AI bids. Of course I’m just an idiot on the internet, not the CEO of a behemoth.

I assumed they meant data governance.

Except that government, at least in the U.S., is not doing their job. This administration doesn't want to regulate AI.

Whether or not it is the government's job is to be regulating a specific thing is not as straightforward of an issue as it may seem.

The charitable entity known as the Mozilla Foundation and the development entity known as the Mozilla Corporation are not the same. Nothing wrong with the foundation doing these things with their spare cash, it literally does not impact Firefox at all.

The concern, I think, is that their spare cash is dwindling and thus financial prudence might be beneficial - especially for those who rely on the core Mozilla propositions like Firefox.

This might be financial prudence of sorts - doesn't something like 80% of their yearly monetary contributions come from Google, particularly for search partnerships? If they are concerned that Google will start paying them less because search has diminishing future returns, diversifying their income sources through investments in AI might be a good idea.

> is that their spare cash is dwindling

They're an NGO, you can just... look this up, it's public.

Sure and my time is limited.

The last I heard was that the Google rev-share agreement was on the skids and they stopped developing projects like Thunderbird and Fake Spotter because they were capital constrained.

If you know otherwise then you're better informed than I am I guess

Mozilla's assets increased again in their most recent financial statement.[1] Investing in AI startups will have high risk however.

[1] https://stateof.mozilla.org/pdf/Mozilla%20Fdn%202024%20-%20A...

That’s one interpretation. Another is that people typically support foundations not simply because they “do good,” but because they advance a specific cause the donor personally values. For example, if I donated to a foundation focused on developing cancer treatments, and that same foundation later shifted its efforts to addressing melting ice caps, I would likely feel frustrated, since that was not the purpose for which I chose to support it, and I don't really care that both actions "do good" in the world.

> people typically support foundations

While I agree with the general point that non-profit donors have a legitimate interest in the use of funds, in the particular case of the Mozilla Foundation, I believe the vast majority of that money is from placement fees Google paid to be the default search engine in Firefox. As the pay-for-placement market has evolved and Firefox's browser share has fallen, this income has also fallen dramatically.

On the general point about non-profit donations, legally non-profit's use of funds are governed by their board of directors and charter, which often are not constrained in ways donors may assume, hence the need for due diligence prior to giving.

Mozilla Foundation's 2024 revenue was 60% investments, 28% program service revenue, 11% contributions and grants.[1]

Most of program service revenue was from Mozilla Corporation. They paid Mozilla Foundation a small part of their revenue for trademark licenses, legal services, and so. And most of Mozilla Corporation's revenue was from Google.

[1] https://stateof.mozilla.org/pdf/Mozilla%20Foundation_Form_99...

Although I haven't looked at the actual reports in a long time, I think that's consistent with my understanding. The Google money (program service revenue) was a much higher percent in the past, creating the $1.4B nest egg. Now the Google money is greatly reduced and the majority of the income is from the nest egg.

I intended the percentages to support your point donations were little of Mozilla Foundation's revenue. But you missed Mozilla Corporation held most assets.

Yes, I didn't take your post as disagreeing. I appreciated the current data since I haven't looked lately.

Mozilla's AI investment strategy is more than Mozilla Foundation spending spare cash.[1]

[1] https://stateof.mozilla.org/ledger/

Reply on news.ycombinator.com