I assume everyone knows the Captain American #1 that this is based on, but just in case.
It seems to apply to the Department of Defense. That's the entire military infrastructure; Army, Navy, etc. That's 2.1 million employees.
Now do Ted Cruz.
I live in a hot climate, so it's really the expense of air conditioning.
Small adjustments to the temperature based on whether or not we're home, pre-cooling versus cooling during the heat of the day, etc. makes a big difference on the bill potentially.
I've seen some scenarios where people were able to save hundreds of dollars a year just by adjusting the timing of systems. The price of electricity can go up and down during the day.
Maybe those cases are outliers and it's actually not worthwhile, but it seems compelling. If I can put a system in place for under $100, that will be at least as good as what I have and possibly a significant improvement, I'm interested in trying it.
So which was the easiest?
Thanks for the tip! Others have mentioned that it's very tricky to find one that works offline without the manufacturer's software (website, login, data collection, etc)
Wow, very detailed! Thanks
Censorship and bias are two different issues.
Censorship is a deliberate choice by the deployment. It comes from a realistic and demonstrated need to limit the misuse of the tool. Consider all the examples of people using early LLMs to generate plans for bombs, Nazi propaganda, revenge p*rn etc. Of course, once you begin to draw that line, you have to debate where the line is, and that falls to the lawyers and publicity departments.
Bias is trickier to deal with because it comes from the bias in the training data. I remember on example where a writer found that it was impossible to get the model to generate a black doctor treating a white patient. Imagine the racist chaos that ensued when they applied an LLM to criminal sentencing.
I am curious about how bias might be deliberately introduced into a model. We have seen the brute force method (eg "answer as though Donald Trump is the greatest American," or whatever). However, if you could really control and fine tune the values directly, then even an "open source" model could be steered. As far as I know, the values are completely dependent on the training data. But it should be theoretically possible to "nudge" those values if you could develop a way to tune it.
Just a reminder, it's very difficult to erase anything from the Fediverse and Lemmy chats are not private.
Also, if you have a headphone jack, you could use a 4-way headphone jack splitter and 4 of those little car audio bluetooth adapters. A little hacky, but it wouldn't break the bank.
There is now an Actually Infuriating community on Lemmy! Post things that are beyond just mildly infuriating. It's only mildly infuriating that someone didn't make this sooner!
There is now an Actually Infuriating community on Lemmy! Post things that are beyond just mildly infuriating. It's only mildly infuriating that someone didn't make this sooner!
view more: next ›
Hey, mod here. You should consider posting this in !newtolemmy@lemmy.ca
I'm removing it from here because this community is for posting news articles.
Welcome!