Slant 3D has some good tips for printing, including standoffs part of the main model that print with it, and are only touching the model at relatively small places (around the diameter of .2mm.)

Having an enclosed chamber heater is a big help. I ponied up for the Bambu X1E, which is like the Bambu X1C except with a chamber heater, Ethernet, and $1000 higher in price, mainly because it could handle ABS with ease. The Bambu H2D also has been a solid performer.

From there, I use ABS support material at interface layers, and do supports manually. Reinforcing is key. The prime tower especially. I add a two layer disk primitive in the slicer as a manual raft to ensure the prime tower does not get knocked over on larger prints.

Disclaimer: I should go Prusa, but at the time, I didn't have the cash for an enclosed XL, and needed multiple printers. It might be good if I just waited until the printers with the INDX system get out there before jumping for a H2C, however, the H2C is a known quantity (with its shortcomings and concerns about lockdown.)

I'm using a Bambu X1E (one of two Bambu printer models that cannot be ordered online), a H2D, and soon a H2C. The filtration system is a lot better than the X1 Carbon, and I also have a secondary air purifier system by it. From the air quality meters I have, with the filtration I'm using, it is in safe limits.

You are right about warping. The H-series has chamber heaters, and the X1E also has that. This combined with an enclosed volume greatly helps with warping, to the point where I've done some obnoxiously tall and skinny printers, using ABS support filament for interface layers (which reduces bed adhesion), and have had some solid results, with negligible, if any warping.

School buses are supposed to flash a yellow set of lights before the red lights come on. I worked with someone who received a big fat ticket about passing a school bus. He showed on his dashcam that the bus was stopped, no lights, but as soon as he went past, the lights came on.

Ticket dismissed. However, it took him a ton of time from work to deal with that.

Sad thing we are already having states pushing to do so. Look at the Flashforge debacle.

This. There are some experimental aircraft that people have, where stuff like 3D printing parts is allowed, but the FAA rep is going to want a lot of documentation on the part and computed/theoretical loads on that part before he gives it a pass.

PEKK, PEEK or ULTEM are what I'd do if I need high temperature stuff. PPS-CF comes close, as I bought my Bambu X1E for that stuff. Tullumer stuff looks promising as well.

For stuff that sits in a car and likely won't warp in the sun, I just go ABS or ASA.

All a company has to say is that they use a third party company that uses various data sources to set the optimal price for the time period, and they are off the hook.

She wouldn't even take a look at me.

USENET coming back as a mainstream communications service may not be a bad thing. I'd like to see it a lot more localized with many local/state/regional groups, because in the past, there was always that specter of the sysadmin banning someone who was naughty from the ISP. Now, that's not really an option. It may even be needed to have quiet moderation on the Big Eight.

I wonder how USENET will combat spam, because it took a lot of stuff bolted onto SMTP to at least slow down the spam problem, and if USENET becomes popular for something than binaries, impersonation and other items will happen often.

We can go down that rabbit hole, but it depends on how tyrannical a government is willing to go. North Korea can slam the screws all the way down to the point of installing client software on devices.

However, can other governments go to those extremes and be effective? Some other states ban VPNs, but one can just use things like SSH to get around that prohibition. This is a cat-and-mouse game that some countries can win... but many cannot, unless they remake the entire Internet.

There are things where AI can't do yet. I remember fast food companies spending trillions to try to make an robot that can do what a short-order chef does, so the food prep area can be unmanned, with some dude coming by to dump off the food stuff. However, this is pretty much impossible, because of cleaning. Same reason why the old vending machines that would dispense sodas, hot chocolate, or chicken soup into a glass or cup are gone. You need people to mop the floors, and ensure nozzles are not fouled.

Trades similar. Good luck having a robot do a HVAC installer's job, or a plumber's job. Even a locksmith. The AI at best may see the lock, but doesn't have any context of what to do next that would be relevant.

The reason is simple... the value of a degree has radically lessened.

In the 60s, a degree from an accredited university, any major, would get someone a job, and job security was high so it pretty was a meal ticket for life, until one did their 20, got their pension.

In the 70s, the decade was in recession, but college with a relevant major could get people work. Worst case, one got their M. S., and then job finding was easy.

In the 80s, things were similar, where one could have a degree in a relevant major. When times went bad, step out, focus on a master's degree, get back in the working world.

Fast forward to these days. A degree means nothing. I've had job interviews turn on knowing the exact specific version of software. Not degrees, but if I knew GolfCart 3.1.3, and no, GolfCart 3.1.2a didn't count. I've had job interviews hinge on certificates, and the degree "requirement" was always tempered by "or equivalent work experience".

The only place a degree really matters is if one wants to become a commissioned officer in an Armed Service branch, or work in the educational sector. Pretty much everywhere else, a degree doesn't matter.

Now, getting a degree for a degree's sake? Definitely something people do, and is what I did, but that is a different thing as opposed to a degree that is an investment for a career.

To me, AI is like the discovery of the electric motor. It can affect everything across our entire life. However, the key is how it is used. When the electric motor was invented, people didn't tout that it would make people turning wrenches obsolete. It was made as a very useful tool for stuff at the time, and was a huge improvement in factories from the old "big steam motor in the basement, with tons of gears and belts...".

We also have had AI for decades. Genetic algorithms, fuzzy logic, etc. It is only relatively recently that many businesses are deciding to just go all out and dump their employees, likely for fast cash for stock buybacks to keep them relevant on the exchanges, due to a crappy economy.

AI is a big thing. However, it isn't perfect. I have had cases where I've asked for a page of code for a specific function... received absolute garbage back with instructions that were not even present. I was better off just writing from scratch. A few months later, I asked the same query, found that it was about 75% of the way there. The problem is that last mile of debugging... and making sure AI is doing defensive programming, comments, and relevant documentation. Vibe programming is nice, but if the code is not readable, nor documented, nor in a format that everything else is in, then it just creates technical debt.

We have the LLMs... the issue is applying the LLMs to things. For example, can a LLM write assembly, hand-tuned, as good, if not better than a dedicated programmer? If so, then it is doing more than decades of compiler wortk. Can we get an engine computer that is generic, and can handle many makes/models/types of engines? That would be useful because many companies are not going to bother making ECUs, just so people would have to buy new, overpriced cars, so having a universal ECU can be a lifesaver.

However, the big thing is that companies have made AI into an enemy, just like with offshore contractors fresh off the boat. Even though AI is useful, people are pushing back on it at all fronts because it isn't being used for better work... it is used so artists, script writers, composers are not needed.

Two AI assistants? Not sure about other stuff, but this sort of reminds me of the insanely hyped game back in the late 1990s, Daikatana, where one had to ensure AI controlled NPCs made it. It didn't exactly help gameplay back then.

There are other items which will compound the problem. Student loan debt. When those debt defaults start happening, those are going to have a major impact, on par with the derivatives back in 2008. Then, there is the fact that people are trying to stuff cryptocurrency in everything, so if BTC really tanks, everyone pays for it, be it peoples' pensions, 401ks, etc.

Finally, there is one thing that people don't realize, which will cause major business failure: The cloud cliff. In the past, if a company had bad times, they would play fast and loose with licensing, not bother renewing support, or ride the high seas to keep the software they need present. These days, with most stuff subscription based and in the cloud, once the company hits the late cut-off date, their data is rendered inaccessible, and they are pretty much out of business for good. This will kill a lot of companies who might have survived.