Once I was told by a mechanic the following: if car companies where the ones liable for fixing your car, the design would look very much different.
Car companies choose what is simple to manufacture, but eventually making it very hard to maintain. If it where not for regulation, they would give you something that has a timer for breaking. That is how little they respect their customers.
Anecdotal (but at this point one of hundreds I've come across): older car had fuel pump accessible through a little sealed hatch. In current car the same pump was integrated into the fuel tank. It went bad. Guess what? You have to remove the whole fuel tank to service it. And that entails taking the exhaust out of the way. Car companies hate you.
> if car companies where the ones liable for fixing your car
Sidenote: this one of the ideas behind Ida Aukens (controversial, and I would say misunderstood) "You will own nothing and be happy": One way to make companies liable is by making them own and maintain what they produce for the entire product cycle, and you only rent these items.
Of course this needs a very healthy and competitive market with effective regulations to prevent eternal enshittification. And maybe it doesn't work when shareholder value is everything that counts.
Fortunately fuel pumps don't fail very often these days and dropping the tank is not a hard job for a mechanic. But I'll admit it's not very fun for a DIYer.
The most idiotic thing I've had to deal with so far was headlight replacement on a Subaru Outback. On the driver's side, you can't get to the headlight from the top. You have to take out the whole front wheel well liner, remove some other things, and then you can GET to the headlight, but it's way back there and you can't even SEE it. So you have to do the job blind and one-handed. I did it ONCE and hated every second of it.
I find that researchers choices of names for the sake of differentiation is more of a barrier than something helpful. Sometimes it feels like I know nothing, but in reality it is the name of the "technique" or phenomena that does not get parsed by my brain.
Things like "Compute-Optimal Sampling" sound just like any other made up gibberish that may or may not exist. Wordings like "memory-centric subsampling", "search based hyper space modeling", "locally induced entropy optimization" don't get parsed. And more often than not after reading such papers, I've come to find out that it is a fancy name for something a toddler knows about. Really disappointing.
I see what you're saying, but I don't think it applies in this case. Correct use of jargon helps domain experts communicate with higher precision, and papers tend to be written by domain experts for consumption by other domain experts.
Of course there are some (possibly many!) papers where jargon is abused to make something sound smarter. Sometimes this can also happen unintentionally.
In this case, "compute-optimal X" is standard terminology used in large-scale ML model design for finding the most optimal tradeoff with regards to compute when trying to achieve X.
Here, the paper is about finding the optimal model size tradeoff when training on LLM-generated synthetic data. Imagine you have a class of LLMs, from small to infinitely large. The larger the LLM, the higher the quality of your synthetic data, but you will also spend more compute to generate this data ("sampling" the data). Smaller LLMs can generate more data with the same compute budget, but at worse quality.
The paper does some experiments to find that in their case, you don't always want the largest possible LLM for synthetic data (as previously thought by many practitioners), instead you can get further by making more calls to a smaller but worse LLM.
Just copy/paste it into chatgpt and ask it to use less jargon or similar.
You are never going to win the jargon battle. It is what it is. People wrap up entire concepts in a few words and hell if they can be bothered writing out the details of the concept over and over again.
Pretty much this. It was a huge disappointment for me to understand that even if I made a revolutionary discovery and doubled the yield of food crops, people would reproduce and the prices (and availability) of food would not change.
I don't think food scarcity/availability/price plays any appreciable role in developed nations birth rates but I'd be willing to consider evidence that it does.
I think price is a factor. It isn’t uncommon for someone to say, “another mouth to feed.”
When adding people to a family, food is probably the biggest non-negotiable cost going up.
I’d also expect to see a strong correlation between family size and the percent of meals eaten at home. Going out to eat quickly becomes unaffordable as family size grows.
That requires line of sight with the target. Most drone usage right now is far beyond line of sight.
You could get around that by having a relay drone at high altitude, but that complicates the system by a lot, and you'd have to somehow defend your relay.
Perhaps they are not. What is really happening is the world is getting more complicated. And if people don't get smarter, or better informed, their capacity to create problems is going to increase.
