The name of “Amazon” carries cachet. It sounds like impressive experience.
The thing is, Amazon famously treats its employees like shit - even the highly paid ones. There’s a sorta Stockholm Syndrome that develops, where you convince yourself that this misery is the cost of “doing great things” or you decide to bail while you can and go someplace that doesn’t suck.
Managers who spend any amount of time at Amazon tend to be the former. They end up with this mindset that if your team isn’t miserable, you just aren’t working hard enough and that having “tough” conversations means that you berate people until they break.
So beware Ex-Amazon managers. They’re not all bad, but a sudden influx is not a good sign.
(In fairness - One of the best managers I ever had was at Amazon. If he left that place, I’d be thrilled to work with him again. Just. Not there.
The hands-down worst manager I ever had was at Amazon as well.)
I’ve been at places where the corporate policies were just buzzword doublespeak and they are a waste of time. Everyone knows it’s bullshit.
I’ve also been at exactly one place where the leadership team gave serious thought to what their goals were for the organization. Then they wrote down a set of goals that were
simple
coherent
actionable And that actually made it easier to do our jobs. When we had to make a decision, we could actually refer to these principles and use them.
It was crazy helpful!
…And then they hired a fuckton of Ex-Amazon managers into high-level roles and they promptly drove away all the best people and replaced the helpful principles with Amazon’s work-or-die philosophy. So I bailed. 😭
Plus one for prazosin. I have a family member with PTSD nightmares. Prazosin has made them able to actually sleep for the first time in their adult life.
Eating super refined sugar like candy causes a chain raccoon of events in your body to occur:
Your blood sugar spikes. Even in non-diabetic people, eating a sugary treat can cause a spike above 150mg/dL (healthy levels are usually below 100 when fasting and below 130 after eating)
Your body ramps up insulin production to help use the sugar.
Because your levels have (probably) spiked so high, your body produces a lot of insulin, but your digestion isn’t sending any more sugar - you already finished digesting it and got the whole load at once.
Your body burns through all the excess sugar in your system and your blood sugar levels fall… precipitously because there’s still a lot of insulin floating around.
Your blood sugar levels fall low enough that your system decides it must be time to eat.
You get the munchies. And since your blood sugar is now lower than your body wants it to be, you crave something extra carby. …Like more candy.
Repeat this process long enough and your pancreas (which makes insulin) burns out. Congratulations! You now (like me) have Type 2 Diabetes! 🎉✨💫
So… To not eat all your candy at once, do a few things to keep your blood sugar from spiking, triggering cravings for more.
Eat fibrous veggies. And eat them at the start of a meal. Getting the fiber into your digestion first slows down processing the rest, keeping your blood sugars from spiking.
Eat carby treats as dessert to a meal, not on an empty stomach. Same reason as above.
If you’ve really gotta indulge, try to drink a Tbsp of vinegar (diluted in a glass of water, so you don’t melt your teeth) just beforehand. The acetic acid in vinegar interferes with digesting carbs. It’s not as good as broccoli, but it helps. (Citric acid does not work like this.)
In addition to the other tips here (Especially keeping treats out of sight), keeping your sugar levels from spiking and crashing helps a lot to curb carb craving.
So, I’ve heard that ML manipulates tokens and specifically for the English corpora they take place of words. If we want model to be polite and not to speak uncomfortable language we can remove certain words from the internal array where all tokens and their associative data are stored, for example “fuck”.
As others have mentioned, it’s not quite that simple.
For starters, you can absolutely remove the word “fuck” from all the training data. Now it’s literally impossible for the AI to “know” the word. But what do you do with the training data? Do you replace “fuck” with a different token? “****” perhaps? Or do you just drop the data entirely?
Giving “offense” is much more complex than just a single word. See, if we just replace the token, the AI may still decide that “Go **** yourself” is a perfectly valid response to a query. On the other hand, if you drop all instances of "fuck"from the data, your AI will just learn offensive euphemisms instead: “You can shove your request where the sun don’t shine”
Worse, there are plenty of sexual / offensive phrases that are built up from perfectly innocuous tokens. “Prone bone”, for instance.
The goal with these (and really almost all) AI models is for them to be “helpful, honest, and harmless”. Simply alerting or replacing a single token (or even combination of tokens) doesn’t really help, because the AI is modeling concepts, not just individual words.
All of this to say that the problem being solved is not to stop an AI from saying “fuck” - it’s to build an AI that doesn’t want to.
What's the purpose of strategy statements and other "corporate plans" in office culture
What exactly is accomplished when corporate culture sits everybody down and has a power point about the strategy, business goals for the next year....
I read that weed makes you not dream, then when you go off weed you have crazy dreams. So heavy, heavy weed users that quit, how long did you have crazy dreams for?
Weeks? Months? Years? Any other interesting experiences?
Which internet meme, if somehow erased from ever emerging, might have the biggest impact?
Is there a fundamental internet meme that has had such a major impact that it would have notable repercussions if it had somehow disappeared?
Do you have a mantra that keeps you going through tough times?
“it do be like that sometimes” is starting to lose it’s magic a little
How do you keep yourself from eating all your candy-stash in one sitting? (lemmy.world)
Many of you have either been trick or treating themselves or have leftover candy from handing it out for halloween....
for ML engineers: why can't you simply exclude the word "fuck"?
So, I’ve heard that ML manipulates tokens and specifically for the English corpora they take place of words. If we want model to be polite and not to speak uncomfortable language we can remove certain words from the internal array where all tokens and their associative data are stored, for example “fuck”.