the_real_sparky

The risks of sudden-stop collisions are better controlled on the track than the street. In addition, traction control has greatly reduced the risk of high-side crashes. Airbags inside the leathers have greatly (almost completely) reduced the risk of torso injuries if you do crash. Helmets are a bit safer than they have ever been (FIM certification). Traction control, ABS and cornering ABS have reduced the risk of dumb crashes while learning, probably by 2/3rds. ABS eventually gets disabled when you get more experienced, but at that point it’s not really needed. Traction control is adjustable on the fly and doesn’t really ever get in the way because of that.

High-level racers have many crashes in their pursuit of faster lap times. The airbags have made an incredible difference in injury rates. I’m just a normal middle aged guy with a family and I find the risks to be acceptable with a fast bike on the track. I understand that most people may not, but then again most people don’t seriously consider the risks they take on a daily basis anyway.

Airbags weren’t fully mandatory in Moto2/3/GP until 2018, but you can see from the 2013-2017 data that it was already safer than you would expect given the high number of crashes. To be fair, some riders were using airbags prior to 2018. Also, crash rates while pushing the edge of the envelope racing are vastly higher than crash rates at track days.

https://www.ncbi.nlm.nih.gov/pmc/articles/PMC8082255/

Define: k = multiple of 1920x1080 pixel count

1920x1080 ~= 2M pixels = 1k 2560x1440 ~= 4M pixels = 2k 3840x2160 ~= 8M pixels = 4k

So now 1440p = 2k, and k becomes meaningful. Problem solved!

It also gives a solution for ultrawides like 7680 x 2160p, which are 8k.

More interestingly, “8K” TVs now become 16k TVs, which marketers should like. We’ve come full circle! Now the k nomenclature also gives you an idea of how difficult that 16k display will be to drive with a video card relative to your existing monitor.

The profit motive for business is to make things easier to use, but ironically that also makes it less likely for anyone using it to have any idea how it actually works. The reason most of us are here is because we had to struggle and learn in order to get computers to work. I don’t think as many kids today will have that opportunity, which probably means that we’ll have more people who use technology but less of them capable of creating it.

I expect that one day I’ll go to a concert and everybody will be wearing them. Hopefully with a synchronized ultra-low latency broadcast option direct from the house mix (hint hint to any Apple employees) to deal with the garbage sound at some areas in arenas.

Adoption of new hardware seems to always be very slow, then all at once.

https://icecalculator.com/home

https://github.com/werman/noise-suppression-for-voice

The Spring Drive is the pinnacle of electromechanical watches so far, and it’s more mechanical than electronic.

“if all seven billion inhabitants on Earth wore a Spring Drive watch, their total power consumption would be just 175 watts”

https://www.ablogtowatch.com/history-seiko-spring-drive-move...

The worst part is that if you just want the package held at the hub in advance then it’s still $5.99 now, even though they didn’t have to actually deliver anything! It’s crazy.

So just wait. I’m sure UPS Norway will soon be coming for your money too.

I do think LLMs will eventually be the solution to this long-standing problem. Even if you could ask questions and it would do a general summary for it and then point you in the right directions for more details we’d be way ahead.

It’s frustrating though as often there are hundreds to thousands of pages of this stuff with diagrams and drawings randomly situated together on the pages. Documentation like this was designed to be dense for printing and consumed by a human that is familiar with it from regular use. I’m a bit concerned that the only solution may be paying a technical expert to sit down and convert it all to blocks of text. It would be an expensive endeavor, and even after it’s complete any changes (which happen often) would have to be continually maintained.

If that’s the only solution then I may still go for it, as I think the value to the business of having all knowledge instantly searchable and then automatically summarized will be considerable.

If you want a real moat, figure out how to parse existing PDF documentation that is really badly formatted. Think diagrams and tables with text floating in various places, etc. Documentation of this style is very common in industries where physical things are being built in the real world. The standards documentation (IEEE, ANSI, NFPA, etc) doesn’t usually parse cleanly, much less the messier internal documentation within the businesses.

Grobid is the best example of such a documentation parser, but it is so laser focused on academic papers that it fails to properly process industry-style standards and SOP documentation. What the world needs right now is a Grobid that works for other kinds of messy documentation.

I have probed it at varying levels of depth in my domain and found it to answer at least as good as an entry-level engineer. This could easily be improved with extended training data. The problem is that when it begins to get things very wrong (or leave out important details), it would be difficult for an inexperienced user to determine that since ChatGPT responds so confidently to every query.

It will be interesting to see how this all plays out.

First, assuming that residential customers can understand the pricing dynamics that are going on is a really bad assumption. It is very complex because there is an interplay between their natural gas service and their electric service, both of which are buying gas from the market on the customer’s behalf. The agents for each company are contractually obligated to supply energy to their customers at any cost - and this is a real problem in this type of situation. In effect, the customer is bidding against themselves via the gas & electric companies. Even in some ideal world where the consumers have full real-time price transparency, it would be extremely difficult for the customer to understand the costs of their actions and the best balance between the two services.

Second, it should be noted that the astronomical gas and electric prices in those markets did not really accomplish anything. They brought on very marginal amounts of additional energy (beyond about 10x average prices there is almost no meaningful additional capacity that can come online because nobody in their right mind would invest in infrastructure that could only operate at those prices). Those prices also did not provide useful information that residential customers could evaluate and act on on the demand side (price sensitive industrial and large commercial customers would shut down at far lower prices than was seen).

Ultimately, what occurred was an emergency situation and disaster. Pricing and consumption controls should have been initiated when energy emergency alerts were declared in the gas and electric markets, as the market structure is not setup to operate effectively during these situations. It was a total failure. The only things that worked were the reliability plans that grid operators have in place to disconnect customers in order to prevent complete failure of the electric grid.

Ultimately, in an emergency/disaster situation there should be a different set of rules that determines who gets the energy and caps the cost.

Note that the winterization discussions going on elsewhere are also extremely important, but those problems are already being covered well.

The best part is that it has almost no impact on voice quality, unlike Krisp and some other options I have tried. Singing into the filter chain even sounds good, with the exception of when my 5 year old daughter joins in. rnnnoise seems to think that her voice is noise and tries to intermittently filter it out, which causes a volume warble while we sing together. To be fair, 99.9% of the time her voice should definitely be considered noise I want filtered out. ;)

https://www.autonews.com/sales/gm-living-large

https://arxiv.org/abs/2002.05534

“ Abnormal respiratory patterns classifier may contribute to large-scale screening of people infected with COVID-19 in an accurate and unobtrusive manner”