There's some code I wrote 30 years ago that's still running inside a text-based, multiplayer, internet game (LPMUD). I wouldn't say it's useful, per se, but it'll probably continue to run for at least several more years.
Yeah, unfortunately, I believe that one side effect of the human (sensory) nervous system having a logarithmic response curve is that humans are bad at understanding exponential processes. If you look at an exponential function on a log scale, you get a graph that looks linear. But, that's what fools people into thinking exponential processes are tame and easy to control.
Well that and it’s hard to distinguish if you’re early in exponential growth or linear. Also exponential growth does hit some kind of ceiling relatively quickly.
I think that’s an easier explanation for why we’re biased against it that isn’t necessarily tied to why the brain uses a logarithmic response curve. The latter could be because it provides better dampening to random excitations (ie the brain doesn’t have to expend as much energy dealing with them at the cost of missing signal in lower energy levels). Of course it could be this is why we’re bad with exponential but that seems like a larger leap to make.
> Well that and it’s hard to distinguish if you’re early in exponential growth or linear.
That's true, but not really saying very much. Any differentiable function is locally linear around a neighborhood of any point where the derivative exists.
> Also exponential growth does hit some kind of ceiling relatively quickly.
Well... that depends. Much like how markets can remain irrational longer than you can remain solvent, exponential growth can often remain exponential for much longer than it takes to create a problem. Conversely, sometimes it can't remain exponential long enough to prevent a problem. Exponential growth is a hard beast to tame.
> That's true, but not really saying very much. Any differentiable function is locally linear around a neighborhood of any point where the derivative exists.
What I’m saying is that it’s hard to know if you are dealing with an exponential scenario. You’d respond differently more quickly but doing so for a linear function may be the wrong response. There are certain failure modes when you do encounter an exponential but most things we encounter are more linear / dampened so the bias humans have against exponential is rational despite the failures we have dealing with exponential problems (climate change being a notable counter example). I’m saying it’s a rational trade off to evolve when dealing with the world.
The word itself suggest as much as well. "Mortgage" is derived from Old French mort ("dead") + gage ("pledge"). In other words, it was a pledge that was supposed to die either when it was paid off, or when the borrower defaulted.
> And it seemeth, that the cause why it is called mortgage is, for that it is doubtful whether the feoffor will pay at the day limited such sum or not: and if he doth not pay, then the land which is put in pledge upon condition for the payment of the money, is taken from him for ever, and so dead to him upon condition, &c. And if he doth pay the money, then the pledge is dead as to the tenant, &c. [Coke upon Littleton, 1664]
If you combine a mortgage with term life insurance matching the remaining principal to be repaid, you can synthetically create that scenario. It’s pretty classic financial advice for a young high single earner starting a family.
I always thought it was interesting that the English word for a home loan is mortgage (meaning death pledge) but the French use another word, hypotheque. I wonder why the French don't use the same word. Maybe the transparency hits too close to home?
20 watts standby is only about 14.7 kWh per month. Even at on-peak, summer PG&E rates, we're only talking about $7/month. Granted, I'd much rather pay $0 than $7, and I'd rather not waste energy, but we're probably not talking about anything close to the amount of energy the average household wastes. I'd be looking at refrigerators and other large appliances for energy savings long before I'd be thinking about how much electricity the TV uses on standby.
That's 20W and $84/yr to do absolutely nothing. There is absolutely, positively no reason the standby power draw needs to be that high. I'm fine with like, 100-1000mW to allow whatever background circuity is required to wake up, but 20W is absurd. That's enough to reasonably illuminate a bedroom (your average 100W-equivalent LED bulb is 14-20W).
$7/month is a really high amount. Especially for nothing. Or something I actively don’t want.
I have lots of dumb devices in my house I don’t want piling on $7 so the manufacturer can try to earn $0.30/month in data fees. I have multiple TVs, a washer, a dryer, a fridge, a cryo tank, lots of things. I don’t want them to waste $7 each.
I am not. I am suggesting there are other savings in the average household energy budget that are more significant. Besides, even "waste" electricity heats a home in the winter, so it's not quite as bad as it sounds even.
I would argue the TV is the most wasteful thing you mentioned. The refrigerator, even if it is inefficient, still does it’s job by cooling food. That ~15kWh monthly to your TV to exist in an off state does nothing for your quality of life. It’s not like the energy is uses somehow charged the TV for future use, it’s just running internal systems that we have no insight into and most likely does not benefit us.
Say you keep the TV for 7 years, that's $588. You can get a 75 inch Sony Bravia TV for $1500. You want to pay 1/3 of the price of your TV just to power it while it is turned off?
Exactly this. More than 7$ a month leave my bank account every month on so many random things, subscriptions I forget to cancel, virtual machines I forget to destroy etc etc. I'd rather have my TV shutdown / turn on quickly over sweating about a random $7 savings on an electric bill that I wouldn't even see anyway.
you don't need to be put into a financial crunch by the $7 in order to see it as wasteful, and it doesn't take 20W of energy to get a television to do things quickly.
we're in this sorry state of affairs because of people accepting lower quality without protest, while expecting everything to be cheap. Companies are in a bad spot, and so are consumers.
Demand doesn't really drive supply, not the way we're told it does. The manufacturers want to make devices that give them a further income stream, and they just have a marketing budget for making sure demand doesn't drop too low for what they're making.
Some feedback from being sort of on the other end here.
I rent a room and the tenant is on the same WiFi network. They politely messaged me one day saying “I think you’re trying to connect to my TV”. Which as it turns out was our toddler mashing the screen on YouTube Kids. I worked out on YouTube kids how to disable the option in the app.
My point being that it’s not always a rational person you’re dealing with, being able to at least disable the option is essential.
By "diversity," do you mean something like "entropy?" Like maybe
H_s(x) := -\sum_{x \in X_s} p(x) log(p(x))
where X_s := all s-grams from the training set? That seems like it would eventually become hard to impossible to actually compute. Even if you could what would it tell you?
Or, wait... are you referring to running such an analysis on the output of the model? Yeah, that might prove interesting....
Because the text we write is not evenly distributed random noise, what we encode into it (by writing) is entropy.
Because LLMs model text with inference, they model all of the entropy that is present.
That would mean that the resulting size would be a measure of entropy (sum of patterns) divided by repetition (recurring patterns). In this count, I would consider each unique token alone an instance of the identity pattern.
While it's not manufactured by the same company anymore, a Korean company bought the formula and sells it on Amazon now! [0] So, you, too, can see what the hype is all about, if you want! At $0.39 per piece of chalk, it is, indeed, expensive. But, it is also a much different experience from writing with ordinary chalk.
> All Hagoromo chalks are well coated to prevent hands from coming into contact with chalk dust when using. Never have to worry about stained, dusty hands when using chalk ever again
You mean not having messy hands after writing on a blackboard is a thing? You mean my whole life was a lie?
https://en.wikipedia.org/wiki/LPMud