The scaling up of battery manufacturing for EVs and now solar storage has lead to prices I would have never imagined I'd see in my lifetime. It's one of the success stories that, having lived through it, has been a real joy.
I know that folks might have been able to point to a graph years ago and said we'd be here eventually, but I had my doubts given the scale required and hacking through all the lobbying efforts we saw against solar/battery. Alas, we made it here!
Alas is right, China is poised to dominate battery, solar, and EV technology and to translate it to military technology as well. Meanwhile the Republicans are blowing up US alliances and sabotaging the battery/EV industrial development policy that was actually making progress in giving the US hope in catching up.
It’s the innovators dilemma. We have so much not just technical but cultural and political sunk cost in fossil fuels and traditional industrial era infrastructure. The Chinese are just developing now and don’t have so much of that sunk cost. So they can think like it’s the future. We are stuck in the past.
Eventually there may come a day when it’s China that is stuck in the past, looking back to the early 21st century like we look back to the middle twentieth, and someone else will be ascendant.
I really felt like Trump’s 2024 election was the moment it became the Chinese century. It was the moment we chose to exit our position of world leadership both culturally and technologically.
Chinese CCP are willing to scarifies whatever traditional industrial era infrastructure in order for things to move forward and gain a global advantage. Especially when they are not the one paying for the scarifies.
Just because a country has previously invested in fossil fuels, it doesn't follow that they can't get the benefit of solar with future investment. However, there's a lot of powerful money/people/corporations that depend on fossil fuels for making billions - that's the real problem as that skews the market and politics of energy production/distribution.
That political sunk cost is why the innovators dilemma happens. It happens in companies too where managers, executives, and top employees will have their careers built around a certain way of doing things. Change threatens that so they will resist change and double down.
Basically success creates the preconditions for this failure mode in the future.
It might be thought of as a form of overfitting. Success results in overfitting to a local maximum.
>I really felt like Trump’s 2024 election was the moment it became the Chinese century.
You must have been asleep at the wheel or living under a rock to have mised China's rise over the last decades. They didn't wait for Trump to get elected in 2024 and then flipped a switch from third world country to global superpower.
"Damn, this hot cup of coffee burned my tongue. Why would Trump do this?"
This is exactly right, IMHO. We were in a course to counter China's momentum, we had handled COVID so much better, our industry had a huuuuuuge investment in it and was poised to take tiff.
And then it was all killed. And we are killing off our other competitive edges over China, the way we attract all the world's best science and tech talent to build here in the US rather than in their own countries. We have sat back scientific research 2-5 years by drastically cutting grants in nonsensical ways and stopping and decimating a class of grad students.
We were the most admired country in the world, and in a short amount of time we have destroyed decades of hard work building a good reputation.
We won't get that back in a year or two, it's going to be decades of work.
This was reported all over, but certain circles considered it politically incorrect to acknowledge that anything good happened in the years 2020-2024, so perhaps you can be excused for missing it. Some random web hits. Check out the graphs herein the massive investment in factories:
Back then when I would inform the politically cloistered about this massive boom in factory construction and the hope for US manufacturing in strategically important energy tech, the most pointed critique was "yeah there's lots of spending but that doesn't mean that the factories are going to make anything." Turns out the skeptics were right. It was a huge mistake that all this stuff went into areas where it is politically incorrect to acknowledge that clean energy is changing the world. Management was not able to trumpet the new investment and the workers dont want to acknowledge what's driving the new higher wages.
As for the US being the most admired country, I work in science and a bit in entrepreneurship. The US was so far and away the leader in these that there's no comparison at all to any other country. Any visitor is completely blown away when they see what's going on, even when they heard ahead of time how much better science and startups are in the US. It's a bit shocking that you think the US was not one of the most admired countries out there, unless you're posting from China or Russia.
It was that Trump and the MAGA crowd conceded to the Chinese by destroying US goodwill and credibility built up over decades. The US will probably never recover those advantages, just as China is ratcheting up its program of dominance. Trump et al have destroyed many things that made the US great.
It's bewildering why anyone would do such a thing but here we are.
Same here in Germany/Europe. Our conservatives actually destroyed the solar industry for the third time. Our conservative party has actually destroyed significantly more jobs in solar industries over the last 20 years than it keeps alive with subsidies of 70k€ - 100k€ per person working in that industry (direct and indirect subsidies make the 70 - 100k€ range).
But hey, our populist right tell us, that the subsidies for "green technology" are bad and that we need to get rid of them, because they are making energy so expensive in Germany (cleared of inflation energy costs are lower than 2013, 12 years ago).
But hey - people vote for those parties. Because they know their economics, not like the leftists, who don't.
Germany (or Europe in general) is fucked. In a few years, we will reap what we now sow. And not because of our social systems or immigration, but because our oh so great political leaders are not willing to invest in the future.
> cleared of inflation energy costs are lower than 2013, 12 years ago
This is not the argument you want to make. Energy prices are a significant component of the basket used to measure inflation. Like yeah, you expect energy prices to sink if you discount for the rise of energy prices. Germany is suffering from high energy prices its the key factor why the country has been stagnating economically for the past 6 years.
Their energy prices are an outcome of incompetence, having tied their energy prices to Russia and a gas supply from them. In hindsight, economic diplomacy is not the path to keeping an authoritarian in check; a strong military and energy independence is.
German energy prices will decline with battery storage and more renewables pushing out the last of their coal and fossil gas generation. Should’ve kept the old nuclear generators running too, as long as possible. Alas, a lesson they’ve learned.
When the iron curtain fell pretty much all of Central Europe liberalized and democratized. The sole exceptions being Belarus and Russia.
Thinking they wouldn’t choose the same path is revisionist.
I remember growing up and gaming online thinking of Russians as nothing strange compared to anyone else. This changed with first Georgia and then very much Crimea.
> An arrangement that began as a peacetime opening to a former foe has turned into an instrument of aggression. Germany is now funding Russia’s war. In the first two months after the start of Russia’s assault on Ukraine, Germany is estimated to have paid nearly €8.3bn for Russian energy – money used by Moscow to prop up the rouble and buy the artillery shells firing at Ukrainian positions in Donetsk. In that time, EU countries are estimated to have paid a total of €39bn for Russian energy, more than double the sum they have given to help Ukraine defend itself. The irony is painful. “For thirty years, Germans lectured Ukrainians about fascism,” the historian Timothy Snyder wrote recently. “When fascism actually arrived, Germans funded it, and Ukrainians died fighting it.”
> When Putin invaded Ukraine in February, Germany faced a particular problem. Its rejection of nuclear power and its transition away from coal meant that Germany had very few alternatives to Russian gas. Berlin has been forced to accept that it was a cataclysmic error to have made itself so dependent on Russian energy – whatever the motives behind it. The foreign minister, Annalena Baerbock, says Germany failed to listen to the warnings from countries that had once suffered under Russia’s occupation, such as Poland and the Baltic states. For Norbert Röttgen, a former environment minister and member of Angela Merkel’s Christian Democrat Union (CDU), the German government bowed to industry forces pressing for cheap gas “all too easily”, while “completely ignoring the geopolitical risks”.
> In February this year, German Green economic affairs and climate action minister Robert Habeck said that gas storage facilities owned by Gazprom in Germany had been “systematically emptied” over the winter, to drive up prices and exert political pressure. It was a staggering admission of Russia’s power to disrupt energy supplies.
> “I was wrong,” the former German finance minister, Wolfgang Schäuble, says, simply. “We were all wrong.”
>> “I was wrong,” the former German finance minister, Wolfgang Schäuble, says, simply. “We were all wrong.”
So when are they retroactively giving back their salaries and pensions for having fucked up the livelihoods of their taxpayers?
Politicians have no incentive to ever make good decisions for the future of their country without any skin in the game from which their personal riches are derived.
>We win or we learn.
Jensen Huang said that failure is learning but sometimes failure is just failure and you should know when to cut your losses before the failure goes from learning to bankruptcy. And Germany did far more failure than actually learning.
> So when are they retroactively giving back their salaries and pensions for having fucked up the livelihoods of their taxpayers?
Does anyone, ever, in any role, do this?
Do CEOs return their bonuses and pay and pensions when they close a business, let alone when they cut the workforce, let alone when they miss the growth of a competitor that is currently still not a direct threat and is instead fighting a battle of attrition with friend of the CEO and would only become a threat if they can take that friend's resources without the attrition destroying everything of value?
> Politicians have no incentive to ever make good decisions for the future of their country without any skin in the game from which their personal riches are derived.
The penalty for most errors in politics is the same as the penalty in any other job: you lose the job.
Most errors, because the really bad errors get you killed, either by an angry mob or by an invading army or by special forces (who may be from the latter while pretending to be the former).
A self inflicted wound.
Europe keeps entering into spot gas supply contracts and paying through the nose instead of signing longer-term contracts for lower prices. The Russians have always been reliable suppliers even after sanctions took place, and calls from some hotheads to use gas as leverage was never seriously followed through by the real decision makers.
And Habeck is an idiot. Lately Germany has not been buying enough summer gas to keep the storage full, and of course the storage gets emptied during the winter - people need to keep warm. To imply that Gazprom is somehow stealing gas from these facilities to exert political pressure is ludicrous, expecially since Gazprom has not even owned these facilities since 2022.
Thanks for the link. I think the facts are correct but the conclusions are wrong. Yamal gas will be redirected to Asian markets by 2030, and Europe will keep losing its manufacturing base to locations with cheaper energy (e.g. the US).
But something tells me von der Leyen will not have trouble heating her own home.
Take a look at US manufacturing activity over the last 12 months. The industry is contracting due to federal policy. US fossil gas prices are rising due to LNG exports, so it is not a sure bet cheap energy is available in the US for manufacturing.
CATL is building one of the largest battery manufacturing facilities in Europe in Spain. I think Europe will adapt without issue to manufacturing without the inexpensive fossil fuels it previously relied on Russia for.
I am not sure how US manufacturing activity contracting implies that Europe is not losing its manufacturing to the US. There are lots of news of European companies expanding in the US (one example would be Airbus in Alabama, lots of others). You are absolutely right about LNG exports, and it's unfortunate because it also pushed residential gas prices up, but just look at the benchmark prices in the US vs. Europe (TTF vs Henry), they are different by a whopping factor of 2 at the moment, and it has been worse in the previous years. Notice that the US manufacturing that tends to concentrate next the the source will get its gas even cheaper.
Volkswagen CEO recently stated that manufacturing in Germany no longer makes sense.
I believe Europe will adapt eventually, but the cost in terms of lost manufacturing and quality of life will be high.
Just look around yourself. No society is comparable to what it was 10 years ago, 20 years ago, etc, in any country. If you think nothing changed in a society, you're just poorly informed.
> cleared of inflation energy costs are lower than 2013, 12 years ago
Dude, soaring energy prices are driving inflation. That's like saying the prices are lower if you just keep ignoring everything that actually makes them more expensive. Duh.
I don’t care if German prices for electricity are below inflation. They’re just still expensive. As an EV owner is difficult to find an electricity provider with costs below 0,25€/kWh, and most of them go beyond 0,30€. While I had prices in other European countries for around 0,05€/kWh at night for example.
Not only that, Conservatives, Socialists and the Green all managed to increase our electricity CO2 footprint by moving from nuclear to coal/lng.
That’s mainly because German has fucked up the smart meter rollout. In their wisdom they separated the meter and the gateway when other countries just combined it. They also made it super secure (good), but then didn’t look at the fact that lots of people live in rented apartments and their meters in the cellars have really poor or no cellular connectivity. When Germany can finally do steerable dynamic loads properly at 95% of the market rather than under 10%, it will finally make a difference on steering pricing for such consumers as yourself.
Germany is investing in massive battery parks dotted around the grid. This will make a difference to supporting base load and offsetting coal, but it will take time.
If there’s anything about the Germans you can count on, is that they move slowly.
Unlikely, since our labor costs are still considerably higher than elsewhere. For a very long time our economy has rested on developing high margin products and letting others do the low-marginal-overhead of making it. We assumed that they were not going to catch up to us as innovators.
That was a dangerous mistake, and we may be left with nothing.
China is governed by the CCP, which holds the world record for the number of people murdered by the state, feeds its citizens militaristic propaganda at scale, is currently controlled by a guy who fancies himself a dictator, and is politically stable only as long as it continues to suppress free speech and free trade. It takes a dangerous kind of willful naivete to just ignore that fact.
I'm sick and tired of whataboutism from people who are somehow motivated to carry water for aggressive dictatorships that threaten the rest of us. I've already lost my birth country to zombies like that (they call them z-patriots, or turbopatriots, the supporters of Russia's invasion of Ukraine). In case you missed it, my original comment was intended as a criticism of the current government of the United States.
This is oversimplified view of the world and China.
China being powerful is not something new, it was the world's largest economy for 18 of the past 20 centuries (with exceptions being parts of the 19th and 20th centuries, when Western Europe and then the US surged ahead after the industrial revolution).
> is politically stable only as long as it continues to suppress free speech and free trade.
Your analysis is through the lens of Western culture. The definition and understanding of freedom and harmony are entirely different in China. I was in China and experienced this myself, so this is firsthand experience, not something I picked up from blogs or news.
In the Chinese context, freedom is defined collectively so freedom from chaos, poverty, foreign domination etc, whereas here in the West it's individual liberty. Harmony and social stability are seen as more valuable than political pluralism, so authoritarian governance is culturally framed as legitimate. You know that 100 million Chinese travel abroad every year and all of them come back to China? Chinese leaders and citizens still remember periods of fragmentation and civil war.
There is a widespread belief that adopting a Western adversarial political model could reintroduce instability and weaken national unity so something China cannot risk given its size and diversity (you know how many ethnicities there are in China?)
This is their natural state. China has a long history of centralized, bureaucratic governance (over 2k years since the Qin Dynasty), where stability and order are prioritized.
Yeah. I have also been to China myself, and have first hand experience walking around Hong Kong with people who later found themselves in jail, or riding the subway getting bombarded with saturation level jingoistic propaganda urging attack against the capitalist aggressors, or getting a tour of Beijing from a friend who worked as a photojournalist and found himself followed by the security services and had to leave and seek asylum with his family.
The silent majority is silent, yes. Those who try to do something get pushed out, or worse. It's the double-edged sword of immigration. But the Chinese people love freedom like the rest of us - you don't need to go far to disprove your entire narrative, Taiwan and Singapore are right there.
Then you've had a very different experience than I have. If you don't mind me asking, where exactly were you in mainland China, and for how long?
Hong Kong isn't representative of China. I've been there and honestly, it felt like a post colonial UK dump. Going directly from Shenzhen to Hong Kong felt like going from a first world country to a third world one, but I digress.
I also talked with Hong Kongers (this year), and they told me a different story, one that isn't so black and white as the worldview you're projecting onto others.
> or getting a tour of Beijing from a friend who worked as a photojournalist and found himself followed by the security services and had to leave and seek asylum with his family.
That's another interesting anecdote. I actually know a photo blogger and a local journalist from China, neither of them is being followed by the security services, and neither has sought asylum anywhere. What was so unique about your friend?
> But the Chinese people love freedom like the rest of us - you don't need to go far to disprove your entire narrative, Taiwan and Singapore are right there.
You know Singapore isn't exactly a "free" country either, right? And Singaporeans are generally fine with that and accept the trade off. So who's disproving whose narrative here?
Different cultures have different systems and trade offs, different value systems and philosophies of life. But some people seem not to understand that and view everything through the lens of their own values, convincing themselves there's only one "right" way to live and that everything else is evil. The Holy Crusades had similar vibes.
That's a funny meta comment, where are you from? Are you consuming a lot of US based content? I ask because I mainly see Americans here writing about the "CCP" based on what they regularly hear from government officials and certain news outlets. It's rarely framed as "China" it's usually "the Chinese Communist Party" emphasizing "Communist" because that word carries negative connotations in the US given its history and in the EU. But maybe framing is similar in your country.
So just to clarify, I'm from the EU, and I'm not paid for anything I write here. Maybe your world model is influenced by propaganda? The world isn't black and white.
I also encourage people to read more about the history and culture of other countries, especially the ones they have strong opinions about, which they often haven't formed themselves (In my experience, this is often lacking in US education, people learn a lot about US history, but not as much about the rest of the world).
Reading more philosophy can also broaden your perspective. In particular, I recommend learning about Singapore, its history, Lee Kuan Yew, and why many highly educated people there willingly accept restrictions on individual freedom. If you understand that, you can then start reading about China, its culture, and its history.
> Are you saying “alas for citizens of the US who see things in competitive nationalist terms”?
He’s saying it as a realist.
China is building the equivalent to America’s sanctions power in their battery dominance. In an electrified economy, shutting off battery and rare earths access isn’t as acutely calamitous as an oil embargo, but it’s similarly shocking as sanctions and tariffs.
Yes and no - yes it’s dumb to give up and let china have a defacto monopoly on the future of energy production. But no insofar as sanctions on battery and solar don’t hit the same as oil and other things. Because once you have them, they keep producing for you.
The change in Chinese output between 2023 and 2024 was an additional 15,000 tons, going from 255,000 to 270,000 tons. The USA's own increased by 3400, from 41,600 to 45,000 tons.
I'm happy to assume Brazilian output will grow, especially if the USA invests a lot in it, but is it going to even be close to enough to make up for where China's already at? China was about 70% of the global output.
Honestly, I don't know. I just know this rare earths business, among other things, was somehow enough for Trump to drop the very deserved Magnitsky sanctions against a brazilian judge.
I hope it was worth it. I have to believe it was. Because otherwise he delegitimized the Magnitsky Act and fucked us in exchange for nothing.
I'm in Berlin, I have more to fear from Trump's administration than from Xi Jinping's.
If I was in the Philippines, I think it would be the other way around. Initially I also had Japan and Taiwan in that comparison, but thinking a bit harder, there's also a risk that Trump is isolationist, that means the risk from each is more like a multiplier than a simple comparison.
For real. I think there's a type of American that would rather hype up the evils of china than admit the distance the US has fallen from its purported ideals. This year I've seen students deported for criticizing Israel, mobs of poorly trained militarized federal police roaming neighborhoods violently disappearing people without trial, the number of homeless grow to 700,000, food kitchens with lines around the block and a president straight up selling pardons to drug dealers.
Chinese totalitarianism just doesn't seem like such a huge contrast as it once did. At least they get an increase in quality of life for the tradeoff. Also a lot of this reeks of Sinophobia tbh
You are certainly not alone in your beliefs, but it always amazes me which technologies get the benefit of doubt and which are severely penalized by unfounded doubt. Solar and especially batteries are completely penalized and doubted in a way that defies any honest assessment of reality. The EIA and IEA forecasts are as terrible as they are because the reflect this unrealistic doubt (random blog spam link, but this observation is so old that it's hard to find the higher quality initial graphs)
Similarly, nuclear power gets way too much benefit of the doubt, which should simply vanish after a small amount of due diligence on construction costs over its history. It's very complex, expensive, high labor, and has none of the traits that let it get cheaper as it scales.
10 new plants at USD 2.7 Billion each. They take six years to build. USD 2/Watt. They have standardised designs, have invested in grownig their manpower and know-how.
If you believe China's internal pricing numbers, sure....
But their actual investments in billions of dollars and in GW show that nuclear is not competing with solar, and is sticking around for hedging bets. They the are deploying far far far more solar and storage than nuclear. And if those nuclear costs were accurate, then nuclear would be far preferable. $2/W is incredible, as in perhaps not credible, but it would also be far cheaper than solar.
And even if China figured out how to build that cheaply, it doesn't mean that highly developed countries will be able to replicate that. Nuclear requires a huge amount of high skill, specialized labor, and doing that cheaply is only possible at certain levels of economic development. As economies develop to ever higher productivity, the cost of labor goes up, and it's likely that nuclear only ever makes sense at a very narrow band of economic development.
In addition to coming so far down in price, it's amazing to me how good the technology has gotten. Batteries that can easily discharge 5C in cold weather, cycle 10000 times, survive harsh conditions with zero maintenance. Panels that last for decades.
Which is why it makes me especially angry that the current US government is throwing away this gift in order to appease a bunch of aging leaders of petro-states. Literally poisoning the world for a 10-15 year giveaway to the richest of the rich.
I take some solace knowing that fossil fuels are now a dead end. And even though certain people are trying to keep the industry going, that end is sooner than ever.
In general it's obvious this is the trend & amazing.
It is a little surprising to me that some markets don't see the benefit. I was pretty delighted ~8 years ago to get some 4500mah 6s batteries RC (under 100Wh) for ~$65 but the price doesn't feel like it's changed much since, based on some light shopping around. Just wanted to note what I perceived as an unevenness. https://rcbattery.com/liperior-4500mah-6s-40c-22-2v-lipo-bat...
My Android phone comes hobbled unless I give it all my data to be used for training data (or whatever). I just asked, "Ok Google, play youtube music." And it responded with, "I cannot play music, including YouTube Music, as that tool is currently disabled based on your preferences. I can help you search for information about artists or songs on YouTube, though.
By the way, to unlock the full functionality of all Apps, enable Gemini Apps Activity."
I'm new to Android, so maybe I can somehow still preserve some privacy and have basic voice commands, but from what I saw, it required me to enable Gemini Apps Activity with a wall of text I had to agree to in order to get a simple command to play some music to work.
When I'm on my bike, it's difficult. I will ride no handed and change a track, but it's more dangerous than it needs to be.
I might switch back to my iOS device, but what I'd really like to do is replace the Andriod OS on this Motorola with a community oriented open source OS. Then I could start working on piping the mic audio to my own STT model and execute commands on the phone.
That seems a lot like corpo- excuse making about adjusting usage to compensate for the fact that a product someone purchased has been changed, and broken, in order to be forced into agreement for a contract. That is called coercion in many places, but it seems like your recommended solution is that people accept getting screwed just so corporations can make more money when people complain…is that correct?
I'm just saying, how often do you need to adjust your phone while you're cycling?
It's a step back to not be able to do it by voice but if you're concerned enough about your privacy, stopping once or twice during a ride doesn't sound like the end of the world.
I'm not saying it's fine that Google took away functionality but, from a practical perspective, it seems like OP was acting like there's no other option available to change tracks. There is and it's really not that inconvenient.
I mean, any time you suggest "regulate companies" or "form a union", you get dogpiled. So until society gets its act together and collectively fixes these problems, the only immediate solution is to opt out.
GrapheneOS or LineageOS on your phone gets rid of the AI cruft. Linux on your computer.
There are few things AI is truly very good at. Surveillance at scale is one of them. Given everything going on in the world these days it's worth considering.
Some parts of voice assistants used AI techniques other parts didn’t. Calling the whole thing AI is like calling Office 365 AI, it’s too vague to be useful. The most reliable parts are using dictation to interact with the preprogrammed bits.
Also, early attempts at dictation wasn’t considered AI, instead a machine learning etc was found to be useful so it’s been tossed into the AI bucket rather arbitrarily.
To be fair it seems to already be happening. My phone keyboard, always prone to interpolating what I type into utter nonsense, seems to have gotten worse in the past year or so.
> Just stop talking to your computer and use the screen interface, that still works.
This reply demonstrates you don't understand the problem. Please don't contribute to the enshittifacation of everything by being an apologist for unethical behavior.
Siri still works fine, I guess. I almost never use it (Android user) but got exasperated with Apple CarPlay's menus and asked it to play something in my wife's car.
Dunno about AI, it works quite a bit like the old Siri, although she's got an iPhone 16 with a current iOS. Worked just fine when I asked it to play some artist other than the annoying YT Music playlists.
Chatgpt is a proprietary eponym[1], like kleenex, or Google for search. That's a relatively strong attractor based on their first mover status. I nevertheless use tissues, and search engines like brave search, sometimes duckduckgo, and claude or openrouter for my LLM models.
I think there are too many good alternatives for Chatgpt to turn the screws too hard on their users, but we'll see where it settles out. As usual, the most vulnerable will be squeezed the hardest (the ignorant and tech feeble). Hopefully competition and some oversight will keep the wolves at bay.
The finance people were chatting about the OpenAI's ad play a while back, glad to see it finally dawning on this crowd.
1. Not all jurisdictions have granted OpenAI the Chatgpt trademark.
Back in '99 Linux didn't run Excel/Word/Powerpoint or most games, but I ran it anyway. What others call showstoppers are for me inconveniences.
I have a motorolla edge 2024 that I'll load whatever open source phone OS will work well enough to place calls and browse the web. I'll keep another phone for the rare times some corporate/government overlord requires it. Many folks who refuse to use smartphones, similarly own a smartphone they rarely use for systems that require them.
My recommendation is to put as little time and energy into closed, locked down platforms as you can. Feel free to complain, but don't forget you can make choices.
Technology has a ratchet effect at scale - as a solution becomes widely adopted, it switches from being a convenience to being a necessity, because people start building more stuff on top of it. It's as true of to-the-minute accurate clocks as it is of smartphone banking.
You can still run a version of Word from 2004. It's fine, if all you need is to write some thoughts down for yourself. But the moment you need to collaborate with other people via a Word document, you'll find it difficult without the modern version with all its user-hostile aspects - and more importantly, other people will find you difficult to work with.
Same applies to other software, web and smartphones, and to everything else in life - the further you deviate from the mainstream, the costlier it is for you. Deviate too much, and you just become a social outcast.
I am nice, it was an illustration of what a logical position/reply would be towards their position "I am a social outcasts". It's a poor argument "it works for me as a social outcast". It's not normal to be one.
Word from 2004 works better than the office 365 version.
I've used it in the last three years to automate document generation in an enterprise because the latest versions of word:
1). Randomly break during automatic updates you can't really turn off.
2). Automatically upload everything to the cloud even when you tell them no.
This isn't the 90s when closed software was better. We are firmly in the enshitification stage of windows and office. Open source is better and is the only sane choice for enterprise.
Those are not words I thought I'd ever write in 2005 or 2015, but here we are.
And we must let someone or some crowd dictate what our basic needs are. That crowd is part of our world. If we stick to our bows and arrows they come with canons and horses. Argh!
That worked fine before agricultural revolution. Since then, if you stick to your bows and arrows, you get sidelined and lose access to benefits of society and civilization.
If it forces you to keep running with more and more speed just to stay where you are, I wouldn't call it as "benefits of society and civilization". A lot of what we call as progress is a forced transformation of basic needs for the gains of business and politics not people.
Even the healthcare, which everyone thinks as a "benefit" of the progress, only resulted in having lopsided demographic pyramid with countries full of old people. I can't think of single scientific result benefiting the human race in its evolutionary goals.
Countries aren't full of old people because of healthcare, they're full of old people because birthrates plummeted after one of the largest generations ever was born in the post-war period.
Causality is complicated and probably impossible to untangle, but the vast decreases in both infant/early child and maternal mortality played a huge role here.
If half your children didn't die by age 20 (or 5), it was possible to have much smaller families. Industrialisation and urbanisation made children net liabilities rather than household assets (providing labour even at a very young age). Financialisation of real estate along with the rest of the economy made earning and saving money critical, and made non-cash or low-cash lifestyles highly marginal (self-sufficient existence or providing many goods and services through the home directly). All that in combination with improved adult lifespans meant that the demographic pyramid consolidated at the bottom and expanded at the top. There are still countries where this isn't the case, most notably now in sub-Saharan Africa, particularly where HIV/AIDS remains endemic:
Interesting to consider this thread with regards to the Amish. They noped off the tech treadmill but it requires a highly cohesive religiously centered society to maintain the necessary critical mass.
It's a lot harder to make an insular society which is self sufficient just to the degree necessary to create an open source smartphone :-p
Technology brings tradeoffs. Conformity in some regards, but it also opens up many new and varied ways of living.
Which is why we need to ban together. Libreoffice isn't dominate, but it has enough market share that it can't be completely ignored. Also if you are using it you are not alone - you are an annoying deviation, but there are enough of you that many cannot ignore you. The more people who also use libreoffice the more power we have. If we can get to just 5% market share we cannot be ignored. (it need not be libreoffice, there are other choices that support that file format well enough which is what we care about.)
LibreOffice's best guess is that they had 200M MAUs in 2019.
I personally find that hard to believe and they don't explain their methodology to arrive at that number (presumably they looked at the downloads and picked a number of users based on feelings).
But, if that number is true, then I suppose you're not only right, but LibreOffice is already near 5% market share.
>but it has enough market share that it can't be completely ignored.
This is the Hacker News bubble in action. Most of the world, most of America, most of China, India, etc. haven't even heard of it. They ignore it and they thrive. Maybe you need to pay attention if you're dealing with certain European governments these days - I'm not sure because I completely ignore it and haven't paid attention since there was just OpenOffice and LibreOffice didn't even exist yet.
> Maybe you need to pay attention if you're dealing with certain European governments these day
Open document formats have been the UK standard for things like .gov.uk for many years. About a decade IIRC. Ignored by some people (notably the Office of National Statistics, of whatever its called these days).
> Most of the world, most of America, most of China, India, etc. haven't even heard of it.
I have come across quite a few non-tech people who use Libre Office.
It has great (some people say better than MS Word with itself between version) compatibility with MS office formats.
I fixed a computer for some old people once who weren't the least bit technical, but they had LibreOffice installed. My guess is they found it searching "microsoft word free" or similar. A bit like how some kids end up finding Minetest/Luanti by searching "free Minecraft".
You can't prove a negative. Usage numbers tell the real story. Either people haven't heard of it, or, worse for proponents, they have heard of it and have decided it's not good enough.
> Feel free to complain, but don't forget you can make choices.
Of course. I can make a choice. When the choice is between being able to login to secure services with my SIM embedded e-signature, use mobile banking and conduct official business and not being able to do any of these things, making choices are easy.
Running Linux on desktop is easy mode when compared to phones, and yes, I started using Linux on desktop in 1999 too with SuSE 6.0. Phones are way more interconnected and central to our lives now when compared to a general purpose computer running your $FAVORITE_OS.
I booted Slackware from a pile of floppies back then. I thought the Germans had a pretty good offering with SuSE at the time.
Look I get it, even back then, most folks felt Windows was the obvious choice (and still do) for their jobs and so on. Sometimes you have to make do with with the unappealing choice in front of you.
For a little more context, my cracked screen iPhone can still do banking or whatever, but I chose not to pony up $800-$1200 for a new iPhone and bought the cheaper $350 Motorolla. It works for me and I think I'm not entirely alone. There are probably some cracked phones, some handme down phones that folks could use for those situations where you really need to use the closed platform, but otherwise are free to use something more open.
Slackware always brings out the inner teen in me. I feel giddy like in the old days. I need to install and maintain it somewhere some time, just for kicks.
I support FOSS wholeheartedly, and believe that it's possible to have a device which is completely Free (not Open but, Free) from hardware design to firmware and software.
On the other hand, there are some nasty realities which bring hard questions.
For example, radios. Radio firmware is something nasty. Give people freedom and you can't believe what you can do with it (Flipper Zero is revolutionary, but even that's a tongue in cheek device). Muck with your airspace and you create a lot of problems. The problem is not technology, but physics. So, unless you prevent things from happening, you can't keep that airspace fair to everybody.
Similar problems are present in pipelines where you need to carry information in a trusted way. In some cases open technology can guarantee this upto a certain point. To cross that point, you need to give your back to hardware. I don't believe there are many hardware security devices with open firmware.
I use MacBooks and iPhones mostly because of the hardware they bring in to the table. I got in these ecosystems knowing what I'm buying into, but I have my personal fleet of Linux desktops and servers, and all the things I develop and publish are Free Software.
I also use Apple devices because I don't want to manage another server esp. in my pocket (because I also manage lots of servers at work, so I want some piece of mind), yet using these devices doesn't change my mind into not supporting Free Software.
At the end, as I commented down there the problem is not the technology itself, but the mindset behind these. We need to change the minds and requirements. The technical changes will follow.
For radios, the general idea of building radios to a spec and having them certified to be sold in country works pretty well most of the time. It might be nice to have a phone with plenty of flexibility on the radio, but I think most folks would be happy just to connect and send work-a-day packets OTA unencumbered by additional restrictions.
It seems like a hardware security device could act similarly to the radio in that the general OS can ask for service (e.g. a signature), but not have access to the internals of the MCU. I don't see why these systems need to be opaque either, in fact it'd be nice to know what is running on the security enclave or LTE radio, even if folks aren't generally meant to access/modify the internals.
It'll be interesting to see how things develop. In my case, I am looking for more experimentation with the smartphone form factor. I'd like to see better options in the market.
I don't think open source and not allowing people to break laws with impunity are at odds. Because there are laws governing airwaves. I think there would need to be some sort of legal entity (foundation?) that would need to steward open firmware + enable it to be locked down so regulations can be followed, but I don't think the two are somehow irreconcilable. The first example that comes to mind is how all the linuxes work with "secure boot" (all of its ridiculousness aside). I think it would be a more effort than that but I truly believe that it is possible to have trust and openness and following regulation. The idea that only a proprietary company can follow the law and comply with regulations is in my opinion strictly false.
That's a big part of the problem: enforcement doesn't scale. It's cheaper to restrict people by legal and technological means, than to let them use judgement and prosecute occasional abusers.
Luckily not everyone agrees with Richard Stallman's hard-line take on proprietary chips.
IMO, if the radio chip just acts as a radio, and passes packets as requested, and any needed firmware blobs are freely distributable, it's fine. It's not ideal, but it's good enough to make a libre-phone.
We all know the network is spying on us anyway, and the radio should be treated as being part of the network, on the other side of the security boundary from the main processor - and since we don't trust it, we don't have to demand that it helps us verify our trust in it!
What about when your smartphone is required to verify your identity so you can work / earn a paycheck? What about when it's required in order for you to engage in commerce?
We're headed down a very slippery slope and the destination is a very dystopian reality where those in power can prevent someone from participating in society on a whim. I believe the destination has previously been described as the beast system or New World Order.
We are all definitely going to have to make a choice. That much is certain.
> What about when your smartphone is required to verify your identity so you can work / earn a paycheck? What about when it's required in order for you to engage in commerce?
In some cases, it already is.
We're already far on the path you described, and there is no choice to make on it, not for individuals. To stop this, we need to somehow make these technologies socially unacceptable. We need to walk back on cybersecurity quite a bit, and it starts with population-wide understanding that there is such thing as too much security, especially when the questions of who is being secured and who is the threat remain conveniently unanswered.
The US is not nearly as far down that path as is, for example, China.
But two forces are at play here:
1. Near-term concern: F-Droid is getting too popular for Google's comfort and Android revenue ambitions
2. Longer term goal: Control. Much of Chinas's social credit scoring is mediated by their phones. Not an issue yet here in the US but assuredly, if not explicitly on the current's government's list of aspirations. A completely managed device with no freedoms (like f-Droid et al,) is antithetical to a more restricted (managed) device.
We're already there. Attestation is not in your phone, but in your ID card. European passports and ID cards carry biometric data of your face, so you can be computationally verified.
I'm aware of this slippery slope for a very long time, esp. with AI (check my comments if you prefer). On the other hand, I believe that we need to choose our battles wisely.
We believe that technology is the cause of these things, it's not. Remember:
Necessity is the mother of invention.
The governments believe that this is the "necessity", so the technologies are developed and deployed. We need to change the beliefs, not the technology.
The same dystopian digital ID allows me to verify my identity to my bank while I'm having my breakfast saving everyone time. That e-sig allows me to have a practical PKI based security in my phone for sensitive things.
Nothing prevents these things from turning against me, except the ideas and beliefs of the people managing these things.
Sure, but the bank feels better about forcing you to interact with their app on a daily basis, because this gives them a direct upsell channel for their financial services. They don't actually want you to us a physical token. Security is only an excuse.
When that security model is based around SIM swappable hardware, this sounds at least questionable. Mobile security seems like a contradiction in itself. I would say this is also why Google is so eager to also lock down the last degree of freedom. So the joke is on you when you use it for online banking
> Feel free to complain, but don't forget you can make choices.
Except, this not really a choice or a reasonable work around.
Phones are still somewhat expensive, not to mention a time-sink to maintain. Try explaining to your parents or even close relatives that they need to abandon the phone they either spent $$$($) on our spend a $$ monthly on that they should really buy another $$$($) phone and use their "official" device like a company card.
Bingo, this right here. Linux desktop wasn’t a daily driver until one day it was.
Although the only problem with this strategy is that Linux got that way because of a lot of private companies that actually wanted that. Valve didn’t want to be locked in with Microsoft. Many of Microsoft’s direct competitors also don’t want to be locked in. IBM famously switched to Mac, Google has been using Mac and Linux workstations for a long time as well.
Also, web technologies like Electron made porting applications to small user bases Linux easier. If that never happened, I wouldn’t be able to use my commercial apps on Linux. This concept might be a little more of a challenge for the mobile app ecosystem, which is a mix of native wrappers like react native and native apps, and there is a high amount of dependency on native APIs for the extra sensors and hardware features phones have the laptops and desktops don’t have.
E.g., For Linux on mobile to work react native can’t be an incomplete implementation like the status quo.
It's a transient state. Food for thought: how much of Linux being a daily driver depends on you having a modern Android or iOS smartphone?
If you need a locked down phone that passes remote attestation to authenticate yourself to a remote service, then whatever you use to access the service UI doesn't really matter: the only device that's necessary to have to use the service is the one you don't fully control, and which gets to control your patterns of use.
An intuition pump I like: imagine you want to put a widget on your desktop that always shows you the current balance of your bank account. You want it to just work ~forever after initial authentication (or at least a couple weeks between any reauth), and otherwise not require any manual interaction. See how hard it is (if it's even possible), and you'll know how badly you're being disempowered already.
Interesting thought. I’d say a low to medium amount but you’re making a good point here.
Most services offer simple SMS two factor, and then if they offer an upgrade to Authenticator or passkey then I have no iOS/Android dependency.
My bank’s website works almost the same as the phone app, I think the only difference is the lack of mobile check deposit (but nobody’s writing checks anymore).
Some services like Venmo are most popular on apps but still have a website.
My remaining hooks are:
- iCloud shared photo libraries with my family. I can use those on iCloud.com but it’s a bit more of a pain. My paid iCloud storage has been migrated to more open alternatives.
- AirTags and Find My. There just isn’t a competitor that’s anywhere near as good. It’s thankfully not a very necessary product.
- Apple Watch. (AirPods actually work great on Linux, btw, even if they are missing some functionality)
- Apple Home. I could migrate this to Home Assistant.
- Apple Wallet. This is mostly convenience. Most things that use it have some kind of alternative, like printed boarding passes. But there’s…
- Ticketmaster. The mobile website tells me I must download the app or add to mobile wallet. Barcodes are dynamic and screenshots don't work. I think the only alternative is to go to the box office before the event which can be very annoying.
My daily driver is Rocky 10, but my control plane is a Pixel 6 on the ATT network but I control almost nothing on that layer. It is why I have been moving most of my core workloads off SaaS and back to local.
Personally I wouldn’t want to have an account with any bank that allowed permanently open api’s - an attacker gets one auth and then can see my balance forever? No thanks.
Yes, I can come up with scenarios where this gives an attacker exactly what they need to time some scam (or mugging) perfectly. I can just as easily come up with scenarios where the same attacker uses already available (or inferrable) information for the same purpose.
Look, many banks are perfectly fine with letting you opt into showing the account balance on their app before log-in step[0]. So why not let someone opt-in to direct access to that information? Or even opt-in to allow the app to expose this information somehow. Even in a body of a goddamn notification[1] (not disabling screenshots is too much to ask, I know, surely everyone will get hacked if this is enabled).
Paranoid mentality about cybersec is a big part of the problem - in itself, but also because it legitimizes the excuses app vendors provide to force users into their monetization funnels.
--
[0] - It's not a very useful feature, since you still need to open the app - and at that point, it's faster to log in via PIN or biometrics than to "swipe down to reveal account balance" or whatever bullshit interaction they gate access through in lieu of just showing the damn thing.
[1] - The increasingly common pattern of "let's notify user that something happened, but do not say what happened in the body of the notification" is getting infuriating. It's another way to force users to "engage" with the app, and it happens to also deny one of the few remaining ways of getting useful data from the app for purposes of end-user automation.
My daily driver has been debian and ubuntu since Potato 25 years ago. My bank has been online only since 2006 and has worked with Konqueror and later Firefox all that time.
2FA is either a standard TOTP generator or an SMS.
Now I do have a smart phone, because I'm not a complete luddite, but I can't think of anything other than perhaps some forms of entertainment (apple tv, paramount, disney perhaps) which might not work on my laptop. I shun things like notifications of my bank balance, is that an essential thing? How did people in the 90s cope without a per-minute balance?
Account balance is a litmus test. If you can't liberate even that information, you've lost control over the banking and your own device.
> 2FA is either a standard TOTP generator or an SMS.
For now. Be grateful while you have it. Most banks everywhere are moving to 2FA through push notifications to their proprietary app, and are deprecating other channels. TOTP is becoming unusual in a bank; where I live, I haven't seen it in use in banking in over a decade (though I'm not counting SMS here; they're technically kind of like TOTP, but they're generated by the service, not on your end).
Between that and a web-wide push for passkeys, having a locked down smartphone is already becoming a soft requirement for doing anything on the web.
"lost control" seems odd, before 1999 I got a bank balance by phoning up a number and putting a ton of other numbers in, so I'm not sure when I ever had control
I guess I could automate my browser or write something, but the lack of a published API doesn't mean I don't have theoretical control over my device (in practice I rely on a linux distribution and firefox/mozilla to create/maintain the browser engine)
Sure in the future they could hypothetically enforce non-free methods to access my bank, and hypothetically all banks could do this, but that's certainly not the case now.
Lots of private companies do not want to be forced to pay Apple and Google a hefty chunk of their earnings either. That's what drove Epic Games and Spotify to fight Apple.
I have a lot of use cases for general purpose computers. If I am operating an event, "inconveniences" are literal showstoppers. When I'm running sound at a performance, switching audio inputs needs to work instantly and with essentially perfect reliability.
Another use case which Linux has a lot of trouble with is operating as a replacement for a pen-and-paper notepad. When I set a computer down for a day, I should be able to turn it on instantly and see the notes that I wrote 3 weeks ago. There are a variety of reasons this doesn't work on Linux. You say "that's an inconvenience" but there are circumstances in which being able to read those notes without needing to wait 30 minutes for the laptop to get enough charge and boot up could be a matter of life or death.
If these kinds of issues are mere inconveniences, that means the computer is a toy rather than a tool.
The problem is as aforementioned players pressure users and government, they can make certain aspects of the economy entirely inaccessible to unapproved platforms. Netflix and co can simply refuse to support streaming on devices which aren't hardware locked. Banks can refuse to do business. Sure banks have in person locations, but they've become fewer and more backed up.
One certain thresholds are reached, little can be done even for the committed outcast.
Well it's true that there's a web option, but it's not the same. It's way more annoying to use IMO (it feels like all your files have to be "in the cloud" ?), and it struggles with big files. On top of that it's less responsive than the desktop version.
> Back in '99 Linux didn't run Excel/Word/Powerpoint or most games, but I ran it anyway. What others call showstoppers are for me inconveniences.
It didn't ran on computer of people that wanted Excel/Word/Powerpoint or most games. I don't think the market of people wanting to use their phone only as a server is big enough for a competitive OS to arise, but I may be mistaken
> it's worth investing a lot of resources in sandboxing.
I tend to agree. There’s an opportunity to make it easy to have Claude be able to test out workflows/software within Debian, RPM, Windows, etc… container and VM sandboxes. This could be helpful for users that want to release code on multiple platforms and help their own training and testing, which they seem to be heavily invested in given all the “How Am I doing?” prompts we’re getting.
Interesting, the pacing seemed very slow when conversing in english, but when I spoke to it in spanish, it sounded much faster. It's really impressive that these models are going to be able to do real time translation and much more.
The Chinese are going to end up owning the AI market if the American labs don't start competing on open weights. Americans may end up in a situation where they have some $1000-2000 device at home with an open Chinese model running on it, if they care about privacy or owning their data. What a turn of events!
sitting here in the US, reading that China is strongly urging the adoption of Linux and pushing for open CPU architectures like RISC-V and also self-hosted open models
It is in their selfish interest to push for open weights.
That's not to say they are being selfish, or to judge in any way the morality of their actions. But because of that incentive, you can't logically infer moral agency in their decision to release open-weights, IP-free CPUs, etc.
Leaving China aside, it's arguably immoral that our leading AI models are closed and concentrated in the hands of billionaires with questionable ethical histories (at best).
I mean China's push for open weights/source/architecture probably has more to do with them wanting legal access to markets than it does with those things being morally superior.
Of course, but that translates in a benefit for most people, even for Americans. In my case (European), I cannot but support the Chinese companies in this respect, as we would be especially in trouble if the common models are the norm.
Depends on whether you want to be in it. A ladder might be enough to peek over the top and rip it off. Do it better. Which seems to be what is happening.
That only works for China's domestic market. As long as the IP they are "taking inspiration from" is protected in the target markets, they effectively lock themselves out by doing that.
In the case of technology like RISC, pretty much all the value add is unprotected, so they can sell those products in the US/EU without issue.
This is exactly what I do. I have two 3090s at home, with Qwen3 on it. This is tied into my Home Assistant install, and I use esp32 devices as voice satellites. It works shockingly well.
I run Home Assistant on an RPi4 and have an ESP32-based Core2 with mic (https://shop.m5stack.com/products/m5stack-core2-esp32-iot-de...), along with a 16GB 4070 Ti Super in an always-on Windows system I only use for occasional gaming and serving media. I'd love to set up something like you have. Can you recommended a starting place, or ideally, a step-by-step tutorial?
I've never set up any AI system. Would you say setting up such a self-hosted AI is at a point now where an AI novice can get an AI system installed and integrated with an existing Home Assistant install in a couple hours?
I mean - the AI itself will help you get all that setup.
Claude code is your friend.
I run proxmox on an old Dell R710 in my closet that hosts my homeassistant (amongst others) VM and then I've setup my "gaming" PC (which hasn't done any gaming in quite some time) to dual boot (Windows or Deb/Proxmox) and just keep it booted into Deb as another proxmox node. That PC also has a 4070 Super that I have setup to passthru to a VM and on that VM I've got various services utilizing the GPU. This includes some that are utilized by my hetzner bare metal servers for things like image/text embeddings as well as local LLM use (though, rather minimal due to VRAM constraints) and some image/video object detection stuff with my security cameras (slowly working on a remote water gun turret to keep the racoons from trying to eat the kittens that stray cats keep having in my driveway/workshop).
Install claude code (or, opencode, it's also good) - use Opus (get the max plan) and give it a directory that it can use as it's working directory (don't open it in ~/Documents and just start doing things) and prompt it with something as simple as this:
"I have an existing home assistant setup at home and I'd like to determine what sort of self-hosted AI I could setup and integrate with that home assistant install - can you help me get started? Please also maintain some notes in .md files in this working directory with those note files named and organized as you see appropriate so that we can share relevant context and information with future sessions. (example: Hardware information, local urls, network layout, etc) If you're unsure of something, ask me questions. Do not perform any destructive actions without first confirming with me."
Plan mode. _ALWAYS_ use plan mode to get the task setup, if there's something about the plan you don't like, say no and give it notes - it will return with a new plan. Eventually agree to the plan when it's right - then work through that plan not in plan mode, but if it gets off the plan, get back in plan mode to get the/a plan set and then again let it go and just steer it in regular mode.
I dont have max plan, but on the Pro i tried for a month, i was able to blow trough my 5 hour limit by a single prompt (with 70k context codebase attached). The idea of paying so much money to get few questions per "workday" seems insane to me
I just wanted to touch on this despite being days later in hopes you see this - I've seen this sort of feedback about the Pro plan quite a bit. I skipped it and went for max so I don't have any experience with it but I can tell you that I've _never hit my/any usage limit_ with the max plan.
Like, I don't know if my account is broken or everyone else just uses things differently. I use claude code, I have it hard-stuck to Opus 4.1 - I don't even touch Sonnet. I _abuse_ the context - I used to /compact early or /clear often depending on the task... but these days (Opus seems much better with nearly full context than Sonnet was) if I'm still on the same task/group of tasks or I think that the current context would be useful for the next thing/task/step I don't even /compact anymore. I've found that if I just run it right up to full and let it auto /compact it does a _really_ good job picking up where it left off. (Which wasn't always the case) Point being - I'm exclusively using Opus 4.1 while also constantly cycling through and maxing out context only to restart with /compact'd context so it's not even starting empty and just keep going.
Hours a day like this. Never hit a limit. (I've said elsewhere that I do believe the general time I work, which is late evening and early morning in north america, does have something to do with this but I don't actually know)
That's great to hear. I was mostly impressed with Qwen3 coder on my 4090, but am hobbled by the small memory footprint of the single card. What motherboard are you using with your 3090s? Like the others, I too am curious about those esp32s and what software you run on them.
Keep up the good hacking - it's been fun to play with this stuff!
I actually am not using the 3090s as one unit. I have Qwen3-30B-A3B as my primary model and it fits on a single GPU, then I have all the TTS/STT on the other GPU.
For the physical hardware I use the esp32-s3-box[1]. The esphome[2] suite has firmware you can flash to make the device work with HomeAssistant automatically. I have an esphome profile[3] I use, but I'm considering switching to this[4] profile instead.
For the actual AI, I basically set up three docker containers: one for speech to text[5], one for text to speech[6], and then ollama[7] for the actual AI. After that it's just a matter of pointing HomeAssistant at the various services, as it has built in support for all of these things.
I assume it's very similar to what Home Assistant's backing commercial entity Nabu Casa sells with the "Home Assistant Voice PE" device, which is also esp32-based. The code is open and uses the esphome framework so it's fairly easy to recreate on custom HW you have laying around.
He is referring the M5 Atom's I believe. I strongly recommend the ESP32 S3 box now, you can fire up Bobbas special firmware for it, search on Github, and its a blast with Home Assistant.
When has the average American ever been willing to spend a $1,000-2,000 premium for privacy-respecting tech? They already save $20-200 to buy IoT cameras which provide all audio and video from inside their home directly to the government without a warrant (Ring vs Reolink/etc).
To be fair, it isn't $1000-2000 extra, it's the new laptop/pc you just bought that is powerful enough (now, or in the near future) to run these open weight models.
Wiredpancake got flagged to death but they’re right. MacWhisper provides a great example of good value for dead-simple user-friendly on-device processing.
You mean like a home with a yard large enough to keep the neighbors out of sight?
Granted, based on how annoyingly chill we are with advertisements and government surveillance, I suppose this desire for privacy never extended beyond the neighbors.
> Americans may end up in a situation where they have some $1000-2000 device at home with an open Chinese model running on it, if they care about privacy or owning their data.
I think HN vastly overestimates the market for something like this. Yes, there are some people who would spend $2,000 to avoid having prompts go to any cloud service.
However, most people don’t care. Paying $20 per month for a ChatGPT subscription is a bargain and they automatically get access to new versions as they come.
I think the at-home self hosting hobby is interesting, but it’s never going to be a mainstream thing.
There is going to be a big market for private AI appliances, in my estimation at least.
Case in point: I give Gmail OAuth access to nobody. I nearly got burned once and I really don’t want my entire domain nuked. But I want to be able to have an LLM do things only LLMs can do with my email.
“Find all emails with ‘autopay’ in the subject from my utility company for the past 12 months, then compare it to the prior year’s data.” GPT-OSS-20b tried its best but got the math obviously wrong. Qwen happily made the tool calls and spat out an accurate report, and even offered to make a CSV for me.
Surely if you can’t trust npm packages or MS to not hand out god tokens to any who asks nicely, you shouldn’t trust a random MCP server with your credentials or your model. So I had Kilocode build my own. For that use case, local models just don’t quite cut it. I loaded $10 into OpenRouter, told it what I wanted, and selected GPT5 because it’s half off this week. 45 minutes, $0.78, and a few manual interventions later I had a working Gmail MCP that is my very own. It gave me some great instructions on how to configure an OAuth app in GCP, and I was able to get it running queries within minutes from my local models.
There is a consumer play for a ~$2499-$5000 box that can run your personal staff of agents on the horizon. We need about one more generation of models and another generation of low-mid inference hardware to make it commercially feasible to turn a profit. It would need to pay for itself easily in the lives of its adopters. Then the mass market could open up. A more obvious path goes through SMBs who care about control and data sovereignty.
If you’re curious, my power bill is up YoY, but there was a rate hike, definitely not my 4090;).
Totally agree on the consumer and SMB play (which is why we're stealthily working on it :). I'm curious what capabilities the next generation of models (and HW) will provide that doesn't exist now. Considering Ryzen 395 / Digits / etc can achieve 40-50+ T/s on capable mid-size models (e.g., OSS120B/Qwen-Next/GLM Air) with some headroom for STT and a lean TTS, I think now is the time to enter but seems to me the 2 key things that are lacking are 1) reliable low-latency multi-modal streaming voice frameworks for STT+STT and 2) reliable fast and secure UI Computer use (without relying on optional accessibility tags/meta).
My greatest concern for local AI solutions like this is the centrality of email and the obvious security concerns surrounding email auth.
Depends on the setup, but programmatic access to a Gmail account that's used for admin purposes would allow for hijacking via key/password exfiltration of anything in the mailbox, sending unattended approvals, and autonomous conversations with third parties that aren't on the lookout for impersonation. In the average case, the address book would probably get scraped and the account would be used to blast spam to the rest of the internet.
Moving further, if the OAuth Token confers access to the rest of a user's Google suite, any information in Drive can be compromised. If the token has broader access to a Google Workspace account, there's room for inspecting, modifying, and destroying important information belonging to multiple users. If it's got admin privileges, a third party can start making changes to the org's configuration at large, sending spam from the domain to tank its reputation while earning a quick buck, or engage in phishing on internal users.
The next step would be racking up bills in Google's Cloud, but that's hopefully locked behind a different token. All the same, a bit of lateral movement goes a long way ;)
I agree the market is niche atm, but I can't help but disagree with your outlook long term. Self hosted models don't have the problems ChatGPT subscribers are facing with models seemingly performing worse over time, they don't need to worry about usage quotas, they don't need to worry about getting locked out of their services, etc.
All of these things have a dark side, though; but it's likely unnecessary for me to elaborate on that.
Given that $2000 might only buy you about 10 date nights with dinner and drinks, the value proposition might actually be pretty good if posterity is not a feature requirement.
The sales case for having LLMs at the edge is to run inference everywhere on everything. Video games won't go to the cloud for every AI call, but they will use on-device models that will run on the next iteration of hardware.
The US is probably ahead but they're so obsessed with moats, IP and safety that their lagginess is self imposed.
China has nothing to lose and everything to gain by releasing stuff openly.
Once China figures put how to make high performance FPGA chips really cheap, its game over for the US. The only power the US has is over GPU supply...and even then its pretty weak.
Not to mention NVIDIA crippling its own country with low VRAM cards. China is taking older cards, stripping the RAM and upgrading other older cards.
Americans may end up in a situation where they have some $1000-2000 device at home with an open Chinese model running on it
Wouldn't worry about that, I'm pretty sure the government is going to ban running Chinese tech in this space sooner or later. And we won't even be able to download it.
Not saying any of the bans will make any kind of sense, but I'm pretty sure they're gonna say this is a "strategic" space. And everything else will follow from there.
When DeepSeek first hit the news, an American senator proposed adding it to ITAR so they could send people to prison for using it. Didn't pass, thankfully.
For criminal concerns regarding retroactive ITAR additions, yes. However, significant civil financial penalties if congress so wished could still be constitutional as the ex post facto clause has been held to apply exclusively to criminal matters starting in Calder v. Bull [1].
History is littered with unconstitutional, enforced laws, as well. Watched a lot of Ken Burns docs this weekend while sick. “The West” has quite a few examples.
There are a lot of things in the US Constitution. But the Supreme Court is the final arbiter, and they're moving closer and closer to "whatever you say, big daddy."
It seems it needs around a $2,500 GPU, do you have one?
I tried Qwen online via its website interface a few months ago, and found it to be very good.
I've run some offline models including Deepseek-R1 70B on CPU (pretty slow, my server has 128 GB of RAM but no GPU) and I'm looking into what kind of setup I would need to run an offline model on GPU myself.
Is there a AI market for open weights? Companies like Alibaba, Tencent, Meta or Microsoft makes a lot sense. They can build on open weights, and not losing values, potentially beneficial for share prices. The only winner is application and cloud providers, I don't see how they can make money from the weights itself to be honest.
I don't know if there is a market for it, but I know that open weights puts pressure on the closed-models companies into releasing their weights and losing their privileged situations.
The only money to be made is in compute, not open weights themselves. What point is a market when a commons like huggingface or modelscope? Alibaba made modelscope to compete with HF, and that's a commons not a market either, if that tells you anything.
By analogy, you can legally charge for copies of your custom Linux distribution, but what's the point when all the others are free?
It promotes an open research environment where external researchers have the opportunity to learn, improve and build. And it keeps the big companies in check, they can't become monopolies or duopolies and increase API prices (as is usually the playbook) if you can get the same quality responses from a smaller provider on OpenRouter
So we're celebrating the real numbers, but maybe we should hoist up the illusory numbers? Back in the day, they thought that some numbers were "imaginary" numbers (e.g. sqrt(-1)) and nowadays, engineers use those imaginary numbers all the time and they feel as real as the reals.
So, here's to math keeping our imagination limber and extending our ideas of what's real.
I know that folks might have been able to point to a graph years ago and said we'd be here eventually, but I had my doubts given the scale required and hacking through all the lobbying efforts we saw against solar/battery. Alas, we made it here!
reply