HN user
I'm not aware of any commercial CAD/CAM software running Linux that Nvidia would have a money interest in supporting. Which current CAD/CAM is running on Linux with graphics? The closest I'm aware of it FEA packages allowing remote calculation on Linux machines while the GUI runs on Windows. Is that what you're referring to?
I'm curious where those number come from. Within the mechanical CAD world where Solidworks is used, I suspect the AutoCAD market share is very close to 0%. I haven't seen any company from small tool shops to major US defense contractors and automotive companies using AutoCAD for any significant mechanical design work.
Yes, all the way up to Version 5.
Canada's policy is to keep the CAD cheaper than USD. No matter how fast USA goes, Bank of Canada can keep up.
Fusion doesn't have the needed tools to large machine designs with multiple components. They have only just added the ability to have assemblies where the parts are separately defined in their own files. Without that, it is only suitable for small, low parts count, one off projects.
We had a sales person try to explain how file handling in 3DExperience is done. They were presenting how magic everything is where the right version of the file is always present in Catia for modelling, Abaqus for FEA, and whereever else in the package.
OK, how do we work with our CMM and CAM software? How do we send files to clients and vendors? crickets
Not confidence inspiring.
I think the main complaint is that the percentage of the kits dedicated to specialty parts limits their usefulness for free-form play. Sets of the 80's had fewer block types, which forced Lego to be more creative in how the sets were put together, which subsequently allowed more freedom in using the blocks for other designs. The sets my kids are playing with look much more "realistic" to what the set is trying to model, but very difficult to build something entirely different, such as building a house from a car kit.
I've heard that many times, but the 3 accounting firms I've worked with for my business didn't care what accounting software I used. They were all happy to work with Gnucash so long as I could provide the needed reports, all of which were pre-configured in Gnucash. Two were small firms, but one was part of a major national accounting firm/franchise.
Are you basing that on viewing what he actually says and does, or through the filter of summaries by people who favour the status quo? Because listening to the new coverage, and then listening to the actual speeches and testimony show opposite conclusions from what I can see.
The adjectives you use seem to be trying to build emotional investment in framing this a good v evil, rather than a sober look at the facts on the ground.
Oh, I think I see what you're saying. If I'm understanding the thrust of your argument:
I do think it would be good if people would be more humble in what they think they know and be more willing to engage with the argument presented by the "other side", and not be so tribal. More introspection, and less blindly doing as they are told, while acknowledging "doctors", "scientists", "reporters", are all actually humans that have human emotions, various incentives, varying knowledge, who sometimes do stupid things, and sometimes things with malevolent intent. They are not all-being, all-seeing, all-virtuous non-humans, so don't take everything at face value.
Many "liberal governments" of the West certainly have some authoritarian elements to them. I don't see that as a conflict with advocating for free speech. If the government is running the propaganda, who is supposed to push against that other than dissidents protected by free speech? It certainly won't be the government or "the authorities".
I don't understand what "YOLO anyone should say whatever and never face rhetorical consequences" means. Who should be enforcing these consequences? What even is a "rhetorical consequence"?
As ever, the problem with creating an authority to regulate what is truth, is who is going to be that authority, and how are we going to prevent it from being corrupted by human nature.
Call it what you will, but the ability of dissenting voices to be heard is the basis of free speech, and also integral to the pursuit of science. Blind trust in authorities is anti-science, and suppression of dissenting views is also anti-science. Those in position of authority like to cast out all who have opposing views as lunatics, but that isn't true. When those in position of authority lie to feather their own nests and cement their power, the truth will be found among the dissidents.
Specifically to Kennedy, in his congressional hearings I've watched does not present himself as a doctor or a scientist, and also not anti-science. His main thrust appears to be that there are a great many problems in the status quo, the "authority" scientists and institutions don't have any reasonable explanations for them, and there are other scientists that are not financially entangled in the status quo that have theories that look to be worth pursuing. That is pro-science in the meaning of exploring the world in pursuit of truth. He is trained as a lawyer, and it is within his profession to be leading inquiries into intent and motivations of various parties in a dispute.
The characterization of him as anti-vax is a slur, and greatly simplified from what he actual advocates.
They don't directly approve drugs, but are involved in the testing pathway for some, and have been caught manipulating data to cause the approval.
https://pmc.ncbi.nlm.nih.gov/articles/PMC2700754/ https://pmc.ncbi.nlm.nih.gov/articles/PMC545012/
Note the dates on those greatly precede Donald Trump ever running for the Presidency.
That is unfortunate, but also, I'd rather choose the situation where truth about abuses of power by authorities can spread with the trade off that some wing nuts are also making up stories out of whole cloth, than the one where truth is crushed under power of authority.
The alternate take is that improved information publishing and distribution platforms (the internet) have allowed the exposure of some pretty corrupt and questionable relationships between the authorities and the industries they regulate (regulatory capture).
Previously people only got their information from the authorities and newspapers. Newspapers were owned by the industries (either directly, or via advertising). Now we can see diverse view points from others in various fields, and it is clear when "doctors say ..." that doesn't mean that all doctors believe that to be true. We can now see that NIH scientists that approve drugs are allowed to approve drugs where they have a patent and commercial interest in the drugs they are approving, which is mind-bendingly wild that level of corruption is allowed.
People can also question where the studies are to back guidelines from authorities. Like what is the scientific basis of the food pyramid? Turns out that was created by the Department of Agriculture to support grain farmers, not because it is a good diet for humans. Or that the deaths and injuries for many infectious diseases had significantly declined before their respective vaccines hit the market, and that the authorities have been cherry picking the points of the graph to hide how much of the improvement happened before vaccines were available.
The biggest change is the availability of diverse voices in an industry being able to be heard, rather than just a select few chosen by "authority", aka power, aka money.
Something is weird there, as the 55+ was increasing in the period of about 1995 to 2010, while the overall was down in point 2000, and 25-54 was flat. Weird.
It seems a bit of a tautology to ask the survivors if things are OK. Of course is is OK for them, they are in an industry segment that survived the trend of moving everything possible to China. If you go ask where the factories were shutdown, you'll find different responses.
The dip in number of manufacturing employees corresponds time-wise to China joining WTO, and the massive CEO hype of shutting down US factories and moving them to China, rather than super-duper automation hitting the scene and automating the factories in place.
I have a suspicion that the displayed manufacturing output chart is in dollars, not amount of stuff, and that is likely bumped up by highly automated IC/chip manufacturing that is high cost/value, but low tonnage work.
While official unemployment is low, that is only people actively looking for work. Labour force participation is way down from it's highs when it peaked in the late 90's. From a peak of around 67%, the US is currently at 62%[1], and the fall off corresponds to the time when imports from China were rising hard and it was the trendy thing to do among executives to shutdown US plants and move production to China.
[1] https://fred.stlouisfed.org/series/CIVPART/
Many people and towns that lost work in the manufacturing shut downs still don't have replacement work, and I think part of that is we lost a lot of low-skilled labour work as the factories left.
Now a question would be, if the work does come back, will it be low skilled enough to be able to hire the same pool of people formerly employed?
I'm split on this. Tradesmen that came up through their apprenticeship say that people need to learn the old manual machines to "feel what the machine is doing", before you can move over to computer controlled machines. But I think that is just the tendency of people to say the path to their current location is the path they took. Not that another path isn't actually a better way to get there for the current environment.
I also think there has been a bit of bias in trade schools towards learning manual processes because it is cheap. You can have a room full of students spend weeks learning to use a file to shape a small block of steel to precise dimensions. A $5 file and $10 of material is easier to supply than a $1XX,XXX computerize manufacturing machine that can process $XXX of material or more per day. But spending weeks to learn how to precision hand file is pretty well a waste of time for modern manufacturing.
I think we're starting to see the pile up that assuming people who worked assembly lines will just become robot tech isn't happening. For the entirety of humans, there have always been a place for low-skilled labour. We had subsistence farming, then as the industrial revolution came on, we ended up with a lot more types of labour jobs. As farms mechanized, we got factories, and as factory output went up, new products were created that needed labour to install them. Now that the remaining factories are requiring smarter workers, farms are pretty well fully automatic, we have a glut of working age people sitting on the sidelines unable to find work that is worthwhile doing.
I think we're at the point where more automation means more loss of work, as opposed to people moving into the new jobs created, for an ever larger portion of the population.
But the current state of a lot of public transport being money losing entities indicates their costs are out of control. Public services shouldn't just help themselves to endless taxpayer money rather than put their house in order and provide value.
I don't think clothing manufacturing is more efficient overseas, it is just cheaper labour. And it is cutting off income stream to lower-skill/intelligence workers in the "advanced" countries. You can see that in the number of working-age men who are no longer in the working population. Not everyone can get a phd and do "advanced, high-value, cutting edge work", any more than everyone could dunk like Michael Jordan if they just trained hard enough.
Also, the financial environment that created "cheap clothes" is also the environment that suppressed middle class wages, and drove the cost of housing up ridiculously. The two are linked. Pretending they aren't is just fantasizing.
Finally, while people are financing larger new houses, the old small houses and apartments are still ripping upward in price. It isn't like people can just choose to pay 1980's prices for 1980's housing stock. They are stuck paying much inflated 2025 pricing for 1980's housing stock.
There could be some legacy places out there that are stuck in the 50's. Also, small businesses seem to be more likely to not have good programs for safety than large outfits. But I've been working in industrial settings for 20 years on 3 continents, and all the big places have had extensive environmental and safety programs. Having said that, I've never been to a steel mill. The workers themselves can be their own worst enemy though. Places that have a large contingent of workers that started in the 70's and 80's tend to think the safety stuff is 'gay', and forced upon them by management.
Have you been to a current US factory? All the big-company ones I've been to have safety and environmental compliance departments focused on zero-injuries and zero-environmental incidents.
Looking at what was done in the early to mid-1900's isn't a good guide to the current state of things. We've learned a bunch since then.
But now both adults need to work to afford a detached house, and the labour participation rate of middle aged men is at an all time low. So, the answer is probably no. I'm sure lots of people would pay and extra $2k for clothing if housing, food, health care, and cars were at similar ratio to wages from the 1980s.
The majority of people are not suitable for management, and do not advance in "careers". Most people just work the same job over and over, until the retire or die.
But the cost of everything has got out of hand. Why bother going to work if you're going to be basically homeless anyway? Housing+food+health care has detached significantly from wages and the costs of other goods.
In the US, housing and real estate have been driven up significantly to the benefit of the really rich. Increased rent, and increased house cost for existing structures just ends up lining the pockets of people rich enough to get priority access to new, cheap debt issuance. Inflation benefits those who are closest to the new money creation, while screwing those furthest away. All the while, everything gets more expensive with increasing rents.
I don't know if that is also the case in China.
We have OpenCASCADE, GCAL and Solvespace. The later two are missing quite a few features relative to OpenCASCADE. Then there is Parasolid, ACIS, and whatever they call the kernel in CATIA. There are a few around. But perhaps there aren't that many ideas for innovative features for people who are interested in the area to create something new.
I think the use of CUDA is niche in the CAD world, probably constrained to a sub-set of FEA. We dropped 6 figures on an Abaqus system -> new hardware, new software within the last couple of years, and a lot of the analysis still runs on the CPU, with a bit of off load to the GPU.
On the performance end, the is a wide range of CAD that was done on systems more than 10x slower, so that also may not be a deal-breaker. For the simpler end of professional CAD operation, I haven't noticed any performance gain from a 2nd get Core i7 to a 10th gen. On the high-complexity end, there is some benefit for some operations, but certainly there is a wide scope of application that could be fine with only 10% of peak performance on newer work stations.
That really seems like it is written by someone without any experience using 3D CAD, particularly of the "make physical stuff" variety, rather than the artistic movie and video game asset variety.
English (or other human) language is pretty bad at describing 3D geometry. People have been using drawings and physical 3D models to describe designs going way back into antiquity. The idea you can just use text suitable for a current-design LLM to manipulate 3D technical data really seems like an idea from someone who has never tried.
But also, Solidworks and Inventor (and I think NX) have extensive API coverage that would allow an LLM to manipulate the geometry data, if the LLM could figure out what the data meant, and what it wants to do. If you can do it on screen with a mouse and keyboard, you can likely do it over the API. An add-in could certainly be written to accept requests for information and instructions on api calls to make from an LLM.