Mozilla Corp., which manages the open-source Firefox browser, announced today that Mitchell Baker is stepping down as CEO to focus on AI and internet safety as chair of the nonprofit foundation. Laura Chambers, a Mozilla board member and entrepreneur with experience at Airbnb, PayPal, and eBay, will step in as interim CEO to run operations until a permanent replacement is found.
Baker, a Silicon Valley pioneer who co-founded the Mozilla Project, says it was her decision to step down as CEO, adding that the move is motivated by a sense of urgency over the current state of the internet and public trust. “We want to offer an alternative for people to have better products,” says Baker, who wants to draw more attention to policies, products and processes to challenge business models built on fueling outrage. “What are the connections between this global malaise and how humans are engaging with each other and technology?”
Diane Brady for Fortune
Mozilla is in such a tough spot. They basically have zero consumer appeal, have no recognisable products other than Firefox, and effectively exist by the grace of Google, of all companies. If Mozilla gets in even more trouble, a lot of OSAlert readers are going to feel it – and the internet as whole will feel the repercussions even if they don’t realise it.
Hearing so much talk about “AI” from Mozilla doesn’t exactly fill me with confidence.
I wonder what’s the percentage of FF users among the Osnews audience?
Dunno about percentages, but I personally use Firefox Nightly on my Android devices. Rarely need Chrome.
cevvalkoala,
*bump*
Thom, browser stats would be relevant and interesting!
Probably higher than the world in general I’m sure, since Firefox tends to get ported to many of the more obscure OSes that OSAlert readers would enjoy using, whereas Chromium is targeted at the three major desktop operating systems (Windows, macOS, and Linux-based).
For me, Firefox at work and Vivaldi at home.
Same here, Firefox (and for certain somethings, Edge) at work, Vivaldi at home.
Not good news, replace bad with bad. Hopefully interim bad will be replaced by good.
I just noticed that it is impossible to close open chat windows within a Facebook profile in Firefox 115.7.0esr. This is an unprecedented situation in the “main” Firefox release. So far, I would rather expect such unpleasant situations in Firefox forks based on the Goanna engine. Tested on a fresh install of Windows 10, a fresh install of the current Firefox ESR, right after the first profile synchronization. Good riddance Mozilla!
The same for me. Things like that will send Firefox to grave. :-/
I’ve been working in tech for decades, and this AI hype cycle is the first bit of “transformative” technology that has me looking forward to retirement. I’ve never felt this way about any new tech before. The hype on this one is stifling.
CaptainN-,
I respect the way you feel about AI, but our opinions about AI, whatever merit they may have, aren’t going to matter much on the macro scale. AI stands to change the job landscape for millions if not billions of people. It’s already begun and the momentum is unlikely to change if big businesses see financial savings.
I mean, that warning narrative is part of the problem. It can’t actually change that much, because all those billions are the same people that drive the economy through their demand. That’s just on basic economics. THe other reality is that these things are not magical, which is my main problem. They are amazing at doing what they are advertised for – processing natural language. They are absolute trash at just about everything else, and while that will improve, it’ll improve across entire sectors, all at once, giving no one any particular advantage. Yes, everyone will be using these to do their work. So what? It used to be no one uses (phones, computers, printers, cell phones, smartphones, the internet, and on and on), now everyone does. Yes, it’s disruptive, it’s transformative, yada yada. But it’s not magic. What will really change at the end of the day? Maybe I’m just getting too old for this crap.
Would you agree that smart phones created more jobs than they displaced? I think things could be very different with AI.
Whilst I of course respect your view, that very much remains to be seen.
PhilPotter,
Sure, but we’re already starting to see it.
https://techcrunch.com/2024/01/09/duolingo-cut-10-of-its-contractor-workforce-as-the-company-embraces-ai/
https://www.cbsnews.com/news/ai-job-losses-artificial-intelligence-challenger-report/
It’s not yet here, but surely coming down the pipeline are AI truck drivers.
Too many people are ill prepared in part because they’re not taking the threat of being displaced seriously.
The vast majority of the layoffs we are seeing now, particularly in the tech sector, are due to post COVID readjustment in my opinion.
As for AI truck drivers and driverless cars that operate in a truly safe and truly autonomous fashion, if 2023 showed us anything it’s that these are further away than ever.
I think we are overestimating just how many jobs are automatable in the short to medium term. Maybe I’m wrong, but I have a strong suspicion AI stocks and firms are in for a drastic downwards value adjustment this year or the next.
Not going anywhere for sure, but a rude awakening is around the corner in terms of actual value propositions for these technologies, until they can be made truly reliable and aligned.
PhilPotter,
Local & city drivers will be the last to be replaced, but highway drivers, it’s only a matter of time. Mark my words, it will start with a few here and there and the trend will accelerate very quickly. Honestly I think the mentality that it’s further away than ever is going to get a lot of those drivers in trouble.
I agree stocks can and will go up and down. They can be overvalued even in a market that has good fundamentals. This doesn’t mean AI isn’t going to displace jobs. One of the main selling points of AI is having it come in cheaper than humans, otherwise what’s the point. However human employees are the greatest expense employers have. And they’re getting more expensive by the year. The economic advantage of AI is only getting stronger. Historically employers have little loyalty to their workers, if they can replace them with cheaper offshore labor, they will and they have and this has displaced millions domestic jobs. I think AI is going to follow a similar trajectory. Maybe it starts with tens of thousands per year, but it will snowball. The reason I am so convinced is because the economics are such that companies are always under pressure to please stockholders, and AI will become the most effective ways to decrease costs and increase profits .
I understand AI needs to become specialized before it can replace a job and it isn’t cheap to develop. However compared to a human, AI scales up insanely well, doesn’t need to sleep or take vacations, etc. And I think it’s going to prove to be more consistent than humans. Even professional jobs like radiologists are at risk of being displaced by cheaper (and maybe even better) AI.
PhilPotter,
PS. I realize there’s no real answer, just predictions. We have to wait and see how it all turns out
Indeed, none of us actually knows for sure. What you’re describing however actually depends on these things existing in a reliable way, and I don’t buy the “AI will keep getting better” line. The evidence seems to suggest it is plateauing already, case in point being how Gemini turned out after all that effort.
The “it’s only a matter of time” line of thought just isn’t supported by the evidence currently (in my view). Also talk of “breakthroughs” and countless arxiv papers, when most of the remotely interesting stuff being done involves a combination of symbolic AI techniques, is premature in my opinion.
Obviously psychologically part of it is that I don’t want to be replaced as a software engineer. That said, I genuinely believe that currently the facts are on my side. Scaling up LLMs by 7 trillion dollars is unlikely to achieve what we want even closely.
PhilPotter,
It sounds premature to say they’re plateauing already. As with any gold rush, you’ll have some wanting to cash in on hype with brand new AI applications may not have much demand and I accept that many of startups will fail. This is normal. However in terms of the business contexts I was referring to, that’s where AI is going to have staying power. These companies already know the demand is there. The real question is NOT about demand, but how to use AI internally in the way that makes the most economic sense for the company. Training AI is expensive today, which creates barriers to entry. But these costs will drop and the bearers will fall. Meanwhile the marginal cost for AI is already unbeatable as it scales extremely well. Human labor remains the highest cost for most businesses and it only gets more expensive.
AI, doesn’t have to be perfect to be competitive against humans, who are notoriously imperfect. When AI comes up in boardrooms, it’s not going to be perfection that sways them, but rather basic economic pressures. Employees will insist they are better than the AI, but for better or for worse profit driven corporations value profits more than employees. They’ll lay off hundreds/thousands of experienced & qualified employees if it means saving a couple bucks here and there. We know they’ll do this because they’ve done it many times before. Entire manufacturing industries have moved continents not because offshore labor was originally better qualified, but simply because it costed way less.
This is exactly how AI gets it’s foot in the door. The first AI will be less experienced than the humans and companies will throw low hanging fruit at it. Over time the roles that AI takes on will increase because the savings are too big to ignore. There will be some holdouts, but after a decade of competition and AI getting better, businesses favoring human labor may end up throwing in the towel.
Just to illustrate this point, Walmart (along with many other corporations) used to have “made in America” campaigns and tried to get consumers to buy US made products. But from high end products to low end products ultimately the cheapest producers won. Now almost all consumer goods are manufactured where it’s cheapest to do so without regards to pride or even quality.
https://www.youtube.com/watch?v=6xxjVZrBeHE
Our corporations are addicted to profits, and this is why AI is going to win.
We shall see
Capitalists are gonna capitalise, that’s for sure, and on that I’m in full agreement. But whether AI in its current or future guises will replace vast swathes of employees depends on many elements. It is just nowhere near good enough now and it really isn’t clear it ever will be, in the foreseeable future at least.
PhilPotter,
IMHO you’re focusing too much on revolutionary AI rather than incremental AI. Many jobs are very close to being displaced by incremental improvements of what we already have. Whether it’s a LLM with more domain specific training needed to do a job, or autonomous drivers, etc. These job disrupters aren’t dependent on some unforeseen technical revolution in AI, but merely the incremental improvements that come with time.
Honestly I think people need to take this more seriously, otherwise society will be very unprepared for the futures. AI is a productivity multiplier and historically this allows employers to increase profits with less labor. I don’t know why people don’t want to take this threat seriously, because it adds up. During the Hollywood writer’s strike, they understood better than most what was at stake and decided to negotiate better terms, but many workers have no representation and it seems very likely that AI will further shift the balance of power towards employers.
It occurs to me that owners & well positioned workers don’t necessarily care that entry level jobs get displaced. But at the macroeconomic level it’s going to be another catalyst for a worsening gap between the rich who reap all the benefits and the poor who get nothing.
I respectfully disagree. You believe incremental improvements of current techniques will displace a lot of work, that’s your right of course.
I on the other hand believe current techniques are fundamentally flawed, particularly in the examples you quote, and it is honestly not clear that incremental improvements will fix them in any meaningful way.
PhilPotter,
The thing is, it’s already happening! We don’t need some unknown future tech, the tech we already have is creating art, writing screenplays, voice acting, driving cars, analyzing medical imagery, translating, etc. These are not perfect, but additional maturity and incremental improvements are going to keep pushing to threaten millions of jobs.
Of course I understand there are other AI milestones to unlock, but it would be a mistake to think one needs to solve more complex AI challenges in order to replace human labor. Human labor is replaceable today, Chatgpt isn’t even specialized. What you’re going to see happening in the coming years is for AI companies to go from one employer to another to train specialized AI models. They don’t have to be perfect to provide value! Companies will use the same performance benchmarks they use to rate human workers. There will be instances where the AI achieves higher performance, fewer mistakes, lower costs. You’ll still have managers but low level jobs will be at risk. Calling AI “fundamentally flawed” doesn’t prevent AI from taking human jobs. AI has no ego and labels aren’t going to stop employers from using AI to maximize their profits. IMHO the most responsible thing for society to do is prepare for this eventuality rather than ignoring it until it’s too late.
You are continually making the same argument. That’s fine, but I still don’t agree with it. I don’t think the evidence shows what you think it does.
I don’t believe the stochastic processes at work here are something we can just incrementally improve. Alignment, genuine understanding, and reasoning skills are just a few of the many problems that will need to be solved to make this genuinely threaten the majority of white collar work. No plausible solutions have yet been posed to any of these issues in terms of how we get from LLMs to what’s next.
I get that you are saying we don’t even need to. I simply don’t agree.
PhilPotter,
You’re trying to portray AI as not being able to take jobs before achieving higher level general AI and/or reasoning, But what is your justification for that? It doesn’t take general AI or reasoning to beat the beat humans at job oriented tasks. Humans being better at matters of philosophy didn’t stop computers from dominating us at specialized intellectual tasks like chess and jeopardy. So now why does AI need a full understanding of the real world or itself in order to be valuable in the eyes of an employer?
I suppose there’s nothing I can say to make you agree, but you have to accept that there are tasks that AI is better at even though it fails at generic reasoning…right? You can agree that, as a productivity amplification tool, AI is able to replace a large team with a much smaller team using AI to increase productivity? Right? To me, these points should not be controversial, and yet I don’t see how we can deny this displaces employees, especially the lower level ones as I’ve been saying.
Yes, I believe it to be (currently) more limited than you do. Perhaps it can make teams of people smaller in limited circumstances, but the economics and results are not there yet, nor is it clear they ever will be with the current LLM scale-up approach.
This is simply me coming to a different conclusion than you have. Comparing chess, a limited and well defined set of challenges, to philosophy, a huge field, is a false dichotomy in my view, and doesn’t prove your point.
I wish you the best though of course, your argument is well thought out, I’m just saying I’m not convinced by it. That isn’t me being willfully blind or burying my head in the sand. It is simply me looking at all the evidence available to me and coming to a particular conclusion based on the sum total of my lived experiences thus far
I may well be wrong.
PhilPotter,
My point was that criticizing AI for lacking broader logic & reasoning skills isn’t a reason AI cannot replace jobs. The are lots of specialized jobs that align with present day AI even though the AI fails at broader logic & reasoning. Anything involving pattern detection is already in the sights of today’s AI models. There are humans doing these types of jobs everywhere from radiology to looking up case laws and I think a lot of these jobs are at risk because AI models can do the job faster, cheaper, and more thoroughly than humans can.
I get we disagree about the future, I was trying to find agreement on the present state of specialized AI, but I guess we won’t get there either. Oh well no matter. I appreciate the discussion!
Thanks, as do I. You are at all times respectful and well argued, and I appreciate that. The last thing I want to do is shout at someone on the interwebs just because they don’t agree with me. That we can disagree and remain polite means a lot to me
Firefox needs to survive until Chrome blocks the efficient ad-blocking extensions. Then people may finally understand and come back to Firefox.
I hope Firefox is not ignored and rendered incompatible by most websites due to lack of marketshare before then.
benoitb,
I’m already seeing some incompatibilities here and there. I hope firefox remains viable, but the shrinking market share is….depressing. It would be a sad day for FOSS alternative browsers if firefox didn’t make it.
Mozilla is like bell labs, they make all this cool stuff… fail to capitalize on it and give it away. They hey kinda half ass thier main product because they are too distracted to FOCUS.
I can’t help but feel like Firefox needs a Mozilla – > Firefox revamp again.
Its trying to force in various things, AI, pocket etc and not really beating Chrome. The new idea is charging people for privacy. Something they could/should be incorporating for free.
Considering they spend $7m a year on the last CEO watching them spiral down, the new one seems to be doing much of the same.