I've sort of been thinking about this as well. Personally, I'd like to re-capture the era of personal blogs and niche knowledge discovery of the earlier web I experienced - crossed with something easy to host/publish and not requiring a browser.
I don't really have any coherent picture but I would like to see these ideas I think:
- Anti-commercial/anti-tracking: maybe requiring some sort of open-source license for all published content that makes it harder to commercially exploit the information, ideally this would be by and for the community, especially in light of recent aggressive LLM-training crawling. I would also like to exclude advertisement and tracking.
- Browser-less: The idea would be to do away with the complexity of the modern web (as people say, browsers are basically operating systems), back to more of its hyper-text roots. Simple documents, mostly textual information. I could imagine a mix of basic markdown and some pre-wired complex/interactive views like "forum" or "blog" and so on (differences in how data is loaded, presented, ...) - the idea would be to implement the "app" part in the browser-replacement and not in the web-page itself if that makes sense. This would lead to more uniformity but that might be a good thing. I'm not even sure if/how images would fit in or videos.
- Peer-to-peer?: Hosting should be as simple as hitting a "publish" button on an article. I like the idea of decentralization, so maybe there could be some sort of peer-to-peer federation where users could "host" content that they've read, liked or general content that's part of a certain (sub-) community. This might require some ranking like HN or a similar mechanism to (unfortunately) censor certain content if the community would not believe it to match their values - so not ultimate freedom. P2P would be more about decentralization, and maybe anti-tracking than pure censorship-resistance.
A session might look like opening the "non-browser" app - it would be fast and require very little memory. Then you'd select or type a community/site and you view of all the content with filters and sorts, depending on the community/site's "template" (again, this is not JS/HTML - basically a native form rendered directly if you will). When you feel like it, you click the "create" button, a text-area + preview pops up and you write your post or article in markdown. When happy, you "publish" and it gets slowly disseminated through all the P2P nodes of your community. This could encompass communities like HN or reddit even if the voting mechanics are worked out, personal blogs, ... but would probably exclude e-commerce stores or video sites because the engine would be potentially too simplistic - and that's fine by me.
Modern smartphones could implement more Data Saver features, but websites could opt-in by using less data. For example, https://marcusb.org/hacks/tinyblog.html
I've been thinking more lately about how to get "Basic Web" - just like normal HTML and maybe a little bit of CSS (No Service Workers, Background Sync, DRM, etc.) and make it work over a LORA/Meshtastic rig somehow.
It's becoming a dream and a curious or nearly unattainable goal, without being adjacent a very heavy software stack. But I think having a separate screen on a phone, like a flip phone with a screen on each side, would help separate the two software and OSes that could run them. One light, and one heavy.
Maybe you're overthinking it. Perhaps instead of creating a whole new stack, you just want to host a "personal blog" yourself. You could even consider creating a listing of similar sites you have found. If all people that want this kind of web host their own personal site I suspect it come to reality. There are of course already several collections of sites like this, 1mb.club 10kb.club nocss.club tilde.town wiby.com sdf.org and more.
You could also consider Gopher or Gemini to find like minded individuals.
But these AIs need energy and GPUs data-centers ... who pays for those? I could imagine a circular mini-economy between a few companies making the bare essentials to keep AI running and not catering to 99.9% of the population because they don't have the funds to buy anything those companies could produce so they don't.
In that scenario AI and robots produce everything, the owners of those AI companies can trade their AI's output with other AI/robot companies, robot and chip manufacturers and commodity owners? So 10000 people world-wide are fine, everyone else is not?
> In that scenario AI and robots produce everything, the owners of those AI companies can trade their AI's output with other AI/robot companies, robot and chip manufacturers and commodity owners? So 10000 people world-wide are fine, everyone else is not?
That scenario is not really any different from having a technology advanced country (like the US) alongside some underdeveloped nation. The US could, in theory, close its borders and produce everything it needs itself, and leave said underdeveloped nation without the benefits of, for example, Netflix, Nike & Nvidia.
But it won’t, it’s not economically efficient. Economic theory (of comparative advantage) tells us that there’s always something that the less developed world can produce relatively cheaply, in exchange for sneakers, streaming services, and G-force Now subscriptions.
That's not why the US is rich. And if the dollar falls, woe to everyone who isn't an American because your future is very bleak in that case. Hope you like digging trenches (and then hiding in them).
Yeah right, it could not. US would probably would be unable to produce something really simple like electrical cables or plastic flip flops, yet alone computer chips.
But on a more serious note, do we know how much Uber spent per technical employee/month? I assume it is far more than even any of those $200 "max ai" plans.
And the other question is how much the public would be willing to spend, in my estimation this is as "cheap" as it will ever get (main-stream at least).
> I assume it is far more than even any of those $200 "max ai" plans.
Am in a random small company, colleague spent 100 EUR a day on Sonnet through AWS Bedrock (needed to use a EU region). Paying for tokens will get you in a deep hole financially compared to any of the subscriptions, unless it's like DeepSeek or one of the other models that are priced a bit better, though that's also a tradeoff in what they can/cannot do and also where the data goes. Ended up trying out the Mistral subscription for the US stuff btw, it was fine.
Exactly. But I did find an article ([1]) and spend doesn't seem that high per engineer ($150 to $250 per eng) - at least on average, I assume the costs were skyrocketing towards the end.
> Adoption climbed from 32 percent of engineers in February to 84 percent classified as agentic coding users by March. By spring, 95 percent of Uber engineers used artificial intelligence tools monthly, and roughly 70 percent of committed code originated from those tools. About 11 percent of live backend updates were written by agents with no human in the loop, according to Uber's own disclosures.
> The numbers behind the spend are what make the story instructive rather than anecdotal. Monthly cost per engineer ranged from $150 to $250 on average, with power users running between $500 and $2,000.
My guess is that the reason to rethink AI-spend was probably the exponential growth in cost over time, and tokenmaxxing payoff not being immediately obvious as mentioned in the article.
Otoh, you can buy contact lenses OTC in other European countries... not saying Europe isn't over-regulated (try legally building a bike-trail on your own forest-land in Germany!! ...insanity) but it's a bad example.
Once you have a reliable printer, the workflow is mostly to slice -> send to printer -> wait and check on it every couple of hours until it's done ime. Imo it no longer super matters how much better the on-screen ui or webcam are.
Mutli-color though is where Bambu has a good leg up.
(Diluted) Vision Miner Nano Polymer Adhesive and a good bed leveling probe has done a lot to make my printer set and forget, no matter which print sheets I use.
Wasn’t the main hassle in calibrarion and Bambu was good in that and is major reason for popularity? So ”once you have a reliable printer” is kinda big thing.
What Prusa is that? Last one I've used (not my own, community lab), I had to level the bed using the sheet-of-paper-method. Which is the reason why I got a Bambu for myself.
My mk3s+ (circa 2020) had a bed sensor. It's still super reliable and I'm only upgrading it to a Core One L because of the upcoming INDX system. Prusa quality, ethics, and true multi-material (not multi-color) in a compact package? Yes please!
Can anyone in the commentariat recommend a great, locally available adhesive in Japan? Vision Miner is import-only and pricey. I’ve been using glue-sticks but am ready to level up as I’m moving away from PLA.
Depending on your level of DIY-ness and willingness to handle powders, you could make some Super Goop. I've heard good things about it but haven't yet had enough bed adhesion issues to make it yet.
I just have a layer of Cape hairspray, on a hardware store aluminium sheet, taped onto the moving down Z frame with Daiso acrylic foam tapes, on a RAMPS1.4 + SERVO42B modified i3-style Cartesian build, works for me.
The motive appears to be to get tax credits as opposed to becoming a full-on patent troll, though with how quickly China is speedrunning their version of capitalism I would not be surprised if it turns into patent trolling.
Their behaviour overall is really giving me mixed feelings, because the Bambu A1 I have is an absolutely amazing machine for the price, and I've been casually in this since the Printrbot days.
If you invent something and publish it (including by offering it to the public in product) your work constitutes prior art and is an absolute bar against the subsequent (valid) patenting of the invention by a third party. F2F vs F2I has no impact on this.
What F2F means if that if two people working in secret create the same invention and show up at the patent office at the same time-- the first one to file gets it. The earlier F2I scheme instead had a contest where the party that is the most ambitious in fabricating lab notebooks to backdate their invention gets the patent.
Because of a poor choice in naming many people wrongfully assume F2F means you can go pick up other people's inventions out of the public sphere and claim them as your own because you filed first.
The misunderstanding is exacerbated because fraudulently patenting other people's inventions is commonplace-- as there is no consequence for doing so except losing the patent after getting defeated on review/litigation-- but the practice isn't meaningfully influenced by F2F vs F2I.
> The earlier F2I scheme instead had a contest where the party that is the most ambitious in fabricating lab notebooks to backdate their invention gets the patent.
The implementation of F2I is then the issue. It should be the first to fulfill all of the following requirements:
(1) physically show at least 2 patent clerks the invention,
(2) that the invention works & operates as outlined, with the clerks being the ones to operate the machine, and
(3) a detailed step-by-step guide to the clerk about how the machine works
The date that the patent is awarded should be the date where the last of the 3 actions occurred.
My thing with bambu was always that they polished whatever the industry (and hobbyists) had invented and closed it all off, then also innovating on top of that but never giving back unless they _had_ to. Polish and mechanical design are great but corexy kinematics, input-shaping are imo what made the X1 stand out as the fast+good-qual printer when it launched. A lot of what they added on top was then to build a moat.
This may be a controversial take, but imo it would be Bambu to set the industry back by a decade if they "win" and lock up the market. That's clearly their strategy afaict.
Does anyone remember Bambu patenting existing open inventions as their own? I can't seem to find good links anymore (?!) but there's some details here https://www.mdpi.com/2411-5134/8/6/141
If no one else is willing to give a polished experience, they have no one to blame but themselves. My father doesn't want to be a 3d printer expert or filament researcher; he wants to print things in 3d as a hobby. Looking back at the reprap, ultimaker, and prusa — the big boys of the maker-oriented printers that i remember — none of them made any progress on making the hobby more accessible to someone like my dad. Bambu deserves some recognition for that.
Prusa is also pursuing patents (they say it's because of Bambu but lol), and they are not releasing their firmware sources for recent printers.
Bambu did not close the tech they used to make their printers. Others (including Prusa) are making CoreXY and they 100% also benefit from the RnD that Bambu did (either hardware, or the slicer (without which Orca would not exist in its current form)).
Bambu just made better products for cheaper and Josef got mad. But I'm certain that Prusa could compete had they focused on making price-competitive polished printers, and not focus on $5k monster printers for enterprise.
I would have said Prusa a year or two ago but they've reneged a little on their open-ness. That was probably in response to Bambu being fully closed and gaining so much market share.
The Core line of printers seems promising and a big leap towards closing the gap towards Bambu's corexy printers but haven't used one yet and I've been out of the game a little. Bambu though is probably more of a high-end appliance type than Prusas more utilitarian feel.
I splurged a while ago and got a pre-assembled Core One. It worked great right out of the box and is has been worry-free so far. So far, I've treated it very much like an appliance with no tinkering on my part yet.
The machine is still quite hackable. Prusa publishes the firmware and CAD files for their printers, although the CAD files aren't under a fully open license. The support is generally nice to people who tinker with their printers and sometimes even seems to be genuinely invested in seeing tinkering projects succeed.
I really like the look of the car, but from the title it sounds like a Mustang has been converted into an FSD Tesla ("teslafied" Mustang) - but Tesla suspension, Tesla interior... this smells like a Mustang body fitted onto a Tesla chassis ("mustangified" Tesla).
I suspect that this might be more of a "Mustang body kit" on a Tesla chassis and not retrofitting the Tesla tech into a Mustang chassis + body. Still cool, but maybe misleading.
Yep, that's exactly what it is. Still a cool project. For a split-second after reading the headline my brain thought they had gotten the Tesla software (with a lot of hackery) to control an ICE vehicle drive-train.
It does work in driving the car where it is able to. Where it fails is in the 'full' part. After 10 billion miles driven on 'auto-pilot' [1] it is hard to claim it 'does not work'. Tesla would have been better off removing the 'F' from 'FSD' but that's water under the bridge.
If we judge it as "Self Driving", it does the Driving part pretty well, and is quite bad at the Self part. There exist no roads and no weather conditions where I'm allowed to take my eyes off the road and stop being the safety nanny.
I liked this article's definition of Full Self Driving (Level 4 autonomy), it is very clear - when Tesla directly takes on the legal liability for unsupervised driving.
You just said "where it fails" and then state hard to claim it not working. If you call it Full Self Driving and it doesn't fully self drive, then it doesn't work. Not really sure where the confusion is. It's not water under the bridge. It is what it is. They claimed it would be fully self driving and not some lane/speed maintenance that pretty much all car makers can do now. It was straight up explained to drive the car. Any deviation from that means it is not working and people like you willingly accepting what Musk has lied about for years trying to make the rest of us out to be the weird ones for not falling for it. I'm tired of the gaslighting.
Literally every non-budget brand (and even some of the budget brands) offers automatic lane keeping with traffic-aware cruise control somewhere in their fleet. It might only be on their flagship vehicles, and possibly only on the top trims, but you're living in a Tesla-decorated cave if you think those are still Tesla-only features.
On a Tesla, it's not even an FSD-specific feature. Autopilot does it.
So many cars come with lane assist and adaptive cruise control. You can google those terms for yourself. I don't bother with lmgtfy requests. You're a big boy/girl, and teaching you to fish it much better effort. They also don't cost an additional $10k on top of the price of the car. They are just part of the price of the car.
Maybe the lmgtfy would be a good exercise for you. Ford's BlueCruise, GM's SuperCruise, Rivian's Autonomy+, and Mercedes-Benz’s Drive Pilot all cost money.
Indeed and none of those work outside of (select) highways. Lucid also has DreamDrive, but it's fairly poor from what I've seen. BYD's God's Eye is in the news, and it isn't looking good either.
I'd love to see good competition in this space, but it seems Tesla has a healthy moat.
I'd love to see waymo adapted to a consumer vehicle, but I have high doubts that this will happen any time soon. I know Waymo has a partnership with Toyota but they don't even have a competitive EV.
Full is a relative term. Full compared to what? Compared to a professional rally car driver? Compared to my grandmother? Compared to a properly licensed tourist in a foreign country?
From videos I see on YouTube, I’m struggling to think what is not Full compared to—at a bare minimum—the bottom 10 percent of drivers on the road.
Yes. If they're restricted they do not have full driving rights. This is of course relative to other drivers. If nobody from your country is allowed to drive in a different country for political reasons or something then not being able to drive there doesn't mean you lack full driving rights. Even someone who has a breathalyzer built into their car doesn't have their full driving rights.
How would you distinguish political reasons from legal reasons? If Bruce is legally precluded from leaving the state of Rhode Island, would we stop describing his brain and body as being capable of the full driving task?
How about if he's allowed to drive anywhere except Rhode Island? Is that any different?
"Full" in full self-driving is a superfluous modifier. But it does is further emphasize that, what a person would consider "self-driving", it can do. Except it can't, of course.
That is brave. Unless you have 5 DWIs to prove how bad a driver you are, or something equally bad... Or maybe you have finally realized old age has destroyed your mind and so you no longer have a drivers license (though this is rare).
Honestly I have no idea how I would objectively rate my driving. I know a few things that I do better than everyone else - but I have no idea what bad things I'm doing that I'm unaware of. I don't know if the bad things I avoid are the really bad things that make me much better, or if they are just minor things and the things I'm unaware of are much more important. About the only thing everyone knows about is that driving drunk is really really bad, but most people don't do that.
Tesla set their own benchmark, their own goal posts, and their own timelines.
In 2016 Tesla said, "as of today, all Tesla vehicles produced in our factory – including Model 3 – will have the hardware needed for full self-driving capability at a safety level substantially greater than that of a human driver.":
Some advice if you want to be intellectually satiated, I would suggest disregarding any forward-looking statements by any company executives. You'll find life to be a lot less emotionally draining.
You can change the subject to the hyperbole and spin by Elon Musk, that's fine. Just be aware that's what you're doing. You're changing the subject. I do understand this cult of anti-personality, but I'm only interested in the technology (made by thousands of people whose name is not Elon) in customer hands right now and not about past promises. Right now, their FSD technology stack looks fairly impressive.
Dude, nobody else is here. You need convince me that I did anything other than ask a reasonable semantic question. Trying to dissemble in front of a non-existent crowd isn't effective. Want to change my mind? Make an actual argument.
Honestly, the best thing to do is go on a free test drive and evaluate for yourself. I don’t see it ever hitting 100%, but I don’t find myself needing to take over under normal conditions.
The steering wheel sensor could be defeated by taping a water bottle to the wheel. The eye tracking is supposed to be harder to defeat, although I wonder if it would accept a mannequin in the driver seat.
I'm at 96% FSD miles since they started tracking a few months ago. It works very well. It was unreliable until about 12-18 months ago, but it's been great since v13.
> The team grafted three sections of the 2024 Tesla Model 3’s floor and seats into the Mustang’s body, shortening the battery case to fit without altering the car’s original dimensions. The result is a classic Mustang shell sitting on top of a Model 3 dual-motor setup
One reason I used it was that I wasn't locked into a single provider and switching them was as easy as changing a drop-down. Small feature? Sonnet or GPT5.4/mini? Large changes? Opus. And why not see how good Raptor Mini does this one refactor?
It also helped build an intuition of what wach model could do and which parts it was weaker at because you could try them almost side by side, especially if one model's output wasn't great.
That said, these were all side projects so nothing truly consequential. Otoh, you might leave some extra perf on the table but I found the models worked quite with the Copilot harness.
Yeah, this is a very useful abstraction layer. The entire concept of separating the model creator from the model runner is good for competition and is customer friendly. Which means they likely hate the concept and want to kill it.
Gosh, imagine getting to do that with your TV/Streaming subscription. Getting to pay one fee to access some set number of hours per month from any of the providers.
I don't really have any coherent picture but I would like to see these ideas I think:
- Anti-commercial/anti-tracking: maybe requiring some sort of open-source license for all published content that makes it harder to commercially exploit the information, ideally this would be by and for the community, especially in light of recent aggressive LLM-training crawling. I would also like to exclude advertisement and tracking.
- Browser-less: The idea would be to do away with the complexity of the modern web (as people say, browsers are basically operating systems), back to more of its hyper-text roots. Simple documents, mostly textual information. I could imagine a mix of basic markdown and some pre-wired complex/interactive views like "forum" or "blog" and so on (differences in how data is loaded, presented, ...) - the idea would be to implement the "app" part in the browser-replacement and not in the web-page itself if that makes sense. This would lead to more uniformity but that might be a good thing. I'm not even sure if/how images would fit in or videos.
- Peer-to-peer?: Hosting should be as simple as hitting a "publish" button on an article. I like the idea of decentralization, so maybe there could be some sort of peer-to-peer federation where users could "host" content that they've read, liked or general content that's part of a certain (sub-) community. This might require some ranking like HN or a similar mechanism to (unfortunately) censor certain content if the community would not believe it to match their values - so not ultimate freedom. P2P would be more about decentralization, and maybe anti-tracking than pure censorship-resistance.
A session might look like opening the "non-browser" app - it would be fast and require very little memory. Then you'd select or type a community/site and you view of all the content with filters and sorts, depending on the community/site's "template" (again, this is not JS/HTML - basically a native form rendered directly if you will). When you feel like it, you click the "create" button, a text-area + preview pops up and you write your post or article in markdown. When happy, you "publish" and it gets slowly disseminated through all the P2P nodes of your community. This could encompass communities like HN or reddit even if the voting mechanics are worked out, personal blogs, ... but would probably exclude e-commerce stores or video sites because the engine would be potentially too simplistic - and that's fine by me.
reply