Mozilla.ai | Hacker News Reader

Mozilla.ai

mozilla.ai

Extensively discussed when first announced six months ago:

Introducing Mozilla.ai: Investing in trustworthy AI (289 comments) https://news.ycombinator.com/item?id=35259817

I love Mozilla's Firefox, but they've got a bad track record of funding things that matter over time. See Thunderbird having to become independent and roll up their own foundation. Mozilla could have had something like OpenAI long ago, but they didn't bother. Instead they bought out Pocket which a lot of loyal users didn't care for. They also stopped maintaining their password manager on iOS (and I'm guessing Android?) which is rather annoying to me, I've been using Firefox's built-in for years, and it was nice to have a PM on iOS. I would have gladly paid money to have it on iOS if it meant it would get funded. There's also their VPN, which is just Mullvad with a skin on it iirc.

Mozilla: It's okay to make products that you sell in order to actually fund your projects properly.

Whoever is managing Firefox (the open source org / corporation, I don't care which) is mismanaging Firefox, and they have been for too long. Mozilla has a lot of potential being burned away. I fear for the day MDN shuts down, since its an invaluable resource when doing front-end web development.

I love and use Firefox every day also, and this is too true.

Bryan Lunduke has a decent skim of some of Mozilla's finances which seems relevant enough to link here: https://lunduke.locals.com/post/4387539/firefox-money-invest...

I wonder what we as a community can do to help preserve the independent browser and web. Mozilla does not seem at all focused to me on this goal. Theoretically it can be changed from the inside?

That's never going to happen. The Mozilla Foundation is a parasite subsisting off of Firefox's name and reputation. The end state is that they eventually kill the host. Google is postponing this to keep their anti-trust shield alive, but it's only a matter of time.

Pretty blackpilled, but this is all rhetoric and opinion, if you’ll forgive me saying so. Do you know anything specific about Mozilla to add?

I don’t know the people in the Mozilla Foundation or its current affairs at all, but I’m hoping others might.

Even if there is no one there who could help change course, could the project in its entirety be “forked” if a new managing body were created and funded?

I also wonder if even MS or Apple, Meta, or even an Opera or other smaller third party with a strong interest in a free web might find reason to join this hypothetical project. Advertisers other than Alphabet?

Kind of reminds me of the sentiment HNers and others have about how Wikipedia spends its funding.

Remember when that was Apple and Microsoft? How times have changed!

> Theoretically it can be changed from the inside?

Eich woulda never let this happen.. At least we got Brave.

We need to be able to fund firefox directly. https://connect.mozilla.org/t5/ideas/ability-to-donate-money...

I'll bite, what's the use case for thunderbird these days now that every email service has its own web and mobile apps?

Because I prefer a seperate client instead of every messaging protocol to rely on a browser or Electron.

Because it's faster to have it auto-open on desktop than to log in (with 2FA) via a browser every time I reboot.

Because I like having a local copy of my email.

Because it's more pleasant to use multiple accounts with a single, consistant UI.

Because I have money for more RAM and compute, but Thunderbird (or other clients) are typically lighter in my experience, so I'll use my RAM or cash for something else.

Aside from people with a single Gmail account (and perhaps 1 work email where they have to use Outlook anyway), who can deal with logging in to multiple webmail apps, all different to each other?

An email client keeps the experience consistent across all accounts. It integrates with whatever workflow and desktop you have. It's cheaper to run. It's easier to use. It's faster. Probably uses less CPU and RAM.

Just another wasted effort. It looks like Mozilla has completely lost its way.

I get that the AI is going to be important in the next decade but Mozilla has shown that despite having a clear competitive advantage in something it cannot continue to grow. I don't mean to devalue what Mozilla is doing but its browser is completely ususable on my Linux Mint 20 with a high end graphics card. Compare that to chrome which works like a breeze.

Although I champion open source & privacy at my org - I cannot but for the sake of my own productivity use any Mozilla product despite repeated attempts every quarter to try if it works better now...

> browser is completely ususable on my Linux Mint 20 with a high end graphics card

Wild. Graphics card model?

Firefox continues to be the most stable/performant browser on all of my devices. Some JS game demos run faster on chromium, yes, but the drop is not significant enough for me to permanently switch.

Same. Even on an old, very low end laptop I have in a drawer, Firefox is more or less the only thing that runs smoothly.

On my MacBook and on my Linux gaming PC with a high-end graphics card (though admittedly not current-gen), it's as snappy as it gets.

What does “unusable” mean? Slow (since you mention graphics card)? Unstable? You don’t like the UI?

> It looks like Mozilla has completely lost its way.

Not at all. $30 million in bonus bumps. The way is looking good!

How is it wasted? So much of their revenue comes from search deals, which are floundering in the courts and being displaced in the wild (by LLM-powered search).

Now, people are paying money to use these LLM products. Which is a niche pricing strategy for search engines.

Nvidia driver? Using latest supported kernel? FF version? Chrome version?

Very vague objective with $30 million thrown at it. That money ain't coming back out.

Haha yeah even with the very specific project of “build a reliable, privacy oriented browser”, you’re scratching your head at where the money went.

Like, “no unsigned extensions” with “oops we forgot to renew the cert, and you can’t opt out, hope your life didn’t depend on any privacy extensions or anything!”

At least they secured mozilla.ai domain

I can't see that as having been of high demand by anyone else.

Scams, likely!

Money well spent haha

Yeah they can host one webpage that links back to Mozilla.org. Nice!

They need to divert profits to stay valid on their non-profit tax exemption. Just a wild guess.

That's not how a 501c3 charity works. There are very strict rules about what the charitable funds can be used for, and "software development" (let alone "AI development") isn't one of those, even if the software is given away for free.

"To establish 501(c)(3) status, your organization needs to be charitable (providing relief to the needy or distressed), educational (improving skills of individuals or providing useful information to the general public), scientific (research in a manner sufficiently in the public interest and not commercial), or in a few other categories (religious, youth amateur sports, etc.)."

https://www.mill.law/blog/more-501c3-rejections-open-source-...

The Mozilla Foundation, the charity part, uses its funds for education.

Funds can be used for operations, which can include software development. (Not really applicable in the Mozilla.ai case, where software would likely be the raison d'etre

Sooo... about the Apache Software Foundation

> To start we will focus on developing tools to build safety [..]

Oh no. Why safety and not pioneering, innovation, or freedom?

Well, I nevertheless wish them all the best.

Right? Like who honestly cares.

I blame all the people who freaked out over making prompts like "create a python function to show me how much to pay employees based on their skin color. "Garbage in, garbage out" should have been the end of discussion there. Instead we have a whole era of useless AIs that give you pages of justification about meaningless "nuance" instead of an answer.

Well, there is lots of low hanging fruit in AI land that, IMO, isn't being taken advantage of.

I'm thinking of stuff like ML compiler integration or other smaller features (including safety features) in the already-popular frameworks being championed by lone hero devs. 2-3 paid ML devs could move mountains right now.

Somehow I'm skeptical, and suspect Mozilla will try to create yet-another-standalone-framework that no one really cares to use, among the pile of 10,000 other from-scratch ML frameworks.

Those low hanging fruit are useless in the face of much more advanced models and even agi coming in the next year or 2. Every effort spent implementing workflows for existing models is pointless when you can just make a better model that has the intelligence to do it automatically.

Not really pointless. I was messing with PlanAndExecute agents for a while, then OpenAI released Functions. Now with Functions I can implement an OpenAI MultiFunction agent that is significantly more reliable, more stable, qualitatively better...

I don't think the efforts that are being expended now to make v0, PoC systems that leverage these newer, more powerful tools is "pointless" - if anything it just helps all participants gauge what is useful, what isn't.

If you want to rely on optimism that AGI is "here" or "around the corner" then sure, go for it... but there's a lot to do in this new paradigm

I dunno about that.

Stable Diffusion is a good example. Some straight up better models, or breaking variants of SD, have come out, but the sheer amount of tooling and augmentations clamped onto Stable Diffusion has kept it dominant.

Llama is another. The intertia and tooling has made MPT and Falcon, and all other non Llama architectures, basically irrelevant.

>much more advanced models and even agi coming in the next year or 2

Are you certain this is going go happen? Especially in such a short timeframe.

What is "trustworthy AI"?

Is that like "accurate AI"?

Or is there some notion of trust I'm missing?

Perhaps 'trustworthy' refers to 'can cite reliable references for all assertions', non-hallucinogenic?

as in:

They "trust me". Dumb fucks.

Or Trusted Computing. Take your pick.

I really don't know what they mean by that. Accurate AI is certainly more specific. As are AI Safety and Privacy. Though Mozilla has botched privacy by putting its encrypted synced data through a confusing UI.

At this point does it make sense to start a new Foundation that moves Blink forward without all the ad/spyware junk that Google and Microsoft distribute with their browsers?

Sponsor the ungoogled-chromium folks, they’re great.

Brave browser, too, though it’s definitely a commercial product.

As far as a foundation though, to me it makes more sense to fund and develop Firefox’s tech, if not that browser any longer.

Love it.

This is exactly what somebody should do. Somebody needs to seriously fork Blink the same way that Google forked Webkit / KHTML way back when.

I'm a fan of Mozilla, but not sure I have a lot of faith in their ability to create anything sustainable; most of their income comes from Google.

> Tools that make generative AI safer and more transparent.

This is great. It quite literally goes without saying that we all understand what this means. It is a boon for us all to celebrate this together now, the moment in which we can all appreciate the scope and details of this project.

This fills my heart with corporate responsibility and legal exposure management emotion.

I'm trying to read between the lines. Their name suggests to me that they're trying to accelerate the development of AI but here they suggest AI safety. I think it's moar AI but with Mozilla's privacy values that so far amount to advertising for Google while keeping the syncing outside of Google, while a lot of users are still drawn to Google with their search partnership.

If they're about AI Safety in the MIRI sense, this doesn't indicate that to me very strongly.

To a layman I would simply propose that one plot a graph of y = x^2 and label the axes as such:

x = mozilla y = safety

less mozilla, more safety? and also more mozilla, more safety?

Yes

" open source AI ecosystem"

Would be incredible, but almost sounds too good to be true. OpenAI and equivalents already have so much money and momentum behind them.

Fully support the idea though!

Somehow I don't trust Harvard's to build something's trustful for the people... Hope I'm wrong ...

Good news. Also, Mozilla should convert to a democratic (one person one vote) worker owned enterprise.

So many people racing to crown themselves as the trusted censor of generative models.

no concrete product or code "We created a company" has such a grifty smell

Fix the url to the root.

This is from at least 6 months ago. Anything new here?

>A startup

You can't claim you are a start up forever.

> Tools that make generative AI safer and more transparent. And, people-centric recommendation systems that don’t misinform or undermine our well-being.

This has nothing to do with Mozilla's core mission. It's verging on abuse of the brand to use it to jump on the AI hype bandwagon like this. I just hope they don't do anything that is toxic to Firefox as it is desperately important that FireFox survive which itself is tenuous enough right now that they should be wholly focused on that in my opinion.

Oh, the negativity.

Look, Mozilla's far from perfect, but I absolutely do want to see them trying something with every new thing that comes out and will continue to throw a little money at them every time they do so.

It's just broadly important to remind the world that there are more ways than "typical capitalism" to do big things in tech.

For me, Mozilla has one very important job, and that's to make a browser. And they are not doing a terrific job right now, they have been losing market share year after year. And sure, you can give me the argument of "but it is the defaults, and ads, and Google, ...", but if they were really good, they wouldn't lose market share, in fact, they took a good chunk of IE back in the days, so much that the joke was that the only purpose of IE was to download Firefox.

So until Firefox become the clearly better browser (like in the IE days) I wish for Mozilla to focus on that and leave AI and other certainly worthy causes to another organization, because the other organizations will not make a browser that is not a chromium fork. Mozilla can get a pass if that's to make money they can invest in the browser, or tech that has direct applications in the making a browser (like Rust). But the AI thing looks like it is neither.

So for me it is money that will not go into fixing bugs in Firefox and as a Firefox user, it annoys me.

I see Mozilla has found yet another thing to focus on instead of building a better web browser.

Great.

</s>

"Mozilla has long championed [AI]..." - paraphrased

Corporations have a straight face saying this when essentially "long" is defined as "since OpenAI's ChatGPT3 blog post on November 30,2022."

https://trends.google.com/trends/explore?date=today%205-y&ge...

The first link shows the Google Trend for 'chatgpt' and the massive spike November 2022. The second link shows the Google Trend of a quiet gentle arc upward in AI until a noticeable spike in Nov 2020 followed by clear explosion ever since

https://blog.mozilla.org/en/mozilla/announcing-the-initial-r...

Reply on news.ycombinator.com