• Xenforo Cloud is upgrading us to version 2.3.8 on Monday February 16th, 2026 at 12:00 AM PST. Expect a temporary downtime during this process. More info here

War Room Lounge V129: Ignored Content Edition

Favorite Chess piece?


  • Total voters
    52
Status
Not open for further replies.
No no Greg, say it with me.

An-Tee-Fer
Douchebags that drive their mom's old BMW with an iPhone 11 and a brand new TV and Apple watch that listened to Dead Kennedys once and now think they are an Anarchist.

So I keep what I was saying earlier.



Truth. Or the difference when people interact with you in more loose social settings like bars. You get a lot of "I'm sorry" for an accident lol.
I'm a big dude, a fatbody if you will.

Working as a bouncer you get SOME people trying to step to you cause they want to prove how tough they are. But the first time regulars saw a fight breakout and me and the other guy that was a former lineman charge the fight and break up a 7 person fight by ourselves suddenly the regulars that might test you... were our best friends.
 
Seriously though what the fuck is wrong with 2020? It’s been rough on a personal level too, anyone else?
I'm actually having a pretty successful year outside all of this insanity

Could lose a few lbs though
 
Seriously though what the fuck is wrong with 2020? It’s been rough on a personal level too, anyone else?

I was supposed to be in Budapest starting May 27th, and I would be in Croatia right now...
 
Seriously though what the fuck is wrong with 2020? It’s been rough on a personal level too, anyone else?

My sister got divorced and had to move in with us for a while. She moved in with a boyfriend within the first month of quarantine which was a Godsend and finally just signed a lease for her own apartment.

Other than that we've been pretty good on the personal front. Our European vacation got canceled but that seems pretty trivial at this point.
 
yA9cK.gif
 
I'm a big dude, a fatbody if you will.

Working as a bouncer you get SOME people trying to step to you cause they want to prove how tough they are. But the first time regulars saw a fight breakout and me and the other guy that was a former lineman charge the fight and break up a 7 person fight by ourselves suddenly the regulars that might test you... were our best friends.

One of my best friends was a bouncer on weekends while in medical school back in the day. He was super jacked, on my best day I couldn't keep up with his workouts. Super hero toned. So during his first few weeks, there were a couple of regulars who would test him. One time a dude went too far, and he punched the guy. One punch broke the guys jaw and KO'd him clean, Hendo/Bisping style of KO. After that night everyone stuck their head so deep that their noses couldn't be any more brown lol. Maybe it's one of those things that you have to do something to "earn the respect" first.
 
I kind of agree with this. Not that I support looting but tbh smashing and burning a few cop cars isn't going to bother me. Is it ideal? No but anytime there's an anti-racist protest people will nitpick it. Block traffic? Oh but what about first responders! Kneel during the anthem? I don't want politics in sports! Why disrespect veterans?!

Its like some people expect protests to happen in the middle of the woods so that they don't bother anyone. Protests have to be disruptive, otherwise they're easily ignored.

My problem is that when you compare these riots to the protest movement of the Civil Rights era the major difference is in organization and discipline. If this was an organized campaign of destruction of law enforcement property I could get behind it. But its just too chaotic, really looks like a minority are there to protest and a bunch of hooligans come along to break and loot shit.
It’s stupid but ok. The best way to effect change in a democracy is to actually vote out the people in office. Get new prosecutors in there, new mayors, new sheriffs. If that’s where the problem is.

If you want to improve policing destroying their assets isn’t going to do it.
 
In machine learning you have 3 main methods: supervised, in which you get training data (which they call features) and associated labels, and the main problem is to classify the labels using the set of features. Then there's unsupervised learning where you just get data, and you try to learn the structure of it without any feedback from a set of labels. And then there is reinforcement learning, where you only get some positive or negative feedback sporadically. The model is that you have an agent, a state, and a set of actions that the agent can take. With each state, action pair, there is a reward associated to it (that reward could be 0). The objective of the agent is to collect as much reward as possible. Older rewards get discounted, which opens up the possibility that, after a bit of searching, the agent would select a policy that is immediately negative, but ends up with a higher total reward after it's all said and done. What is then tricky is typically how to search the state action pairs, which is done with a variety of stochastic approaches (random sampling, stochastic gradient, simulated annealing etc.).
I clicked on the link and read that. Sorry I wasn't clear. Please elaborate on the connection between this and my post--particularly, was that, in fact, the post in that thread you were referring to, and, are you trying to say people who are actual experts in this field are way ahead of me and this their approach to that problem? Has it been proved to work?

If so, that could also be applicable to Bayesian Confirmation Theory if I understand both well enough.
 
One of my best friends was a bouncer on weekends while in medical school back in the day. He was super jacked, on my best day I couldn't keep up with his workouts. Super hero toned. So during his first few weeks, there were a couple of regulars who would test him. One time a dude went too far, and he punched the guy. One punch broke the guys jaw and KO'd him clean, Hendo/Bisping style of KO. After that night everyone stuck their head so deep that their noses couldn't be any more brown lol. Maybe it's one of those things that you have to do something to "earn the respect" first.
That line of work there is 100% an aspect of needing to earn patrons respect.
 
I clicked on the link and read that. Sorry I wasn't clear. Please elaborate on the connection between this and my post--particularly, was that, in fact, the post in that thread you were referring to, and, are you trying to say people who are actual experts in this field are way ahead of me and this their approach to that problem? Has it been proved to work?

If so, that could also be applicable to Bayesian Confirmation Theory if I understand both well enough.

I'm just saying that this is the major approach to the problem, and highlighting some methods that are used to tackle it. As you say in your original post, the main problem is that you don't ultimately know the distribution of the rewards, you have to estimate it, and you don't know when you're done estimating. Which is where things like monte carlo or stochastic gradient come up. This is the approach that AlphaGo takes, so yes it has been ''proven'' to work. In that, machine learning works in a domain of difficult and intractible problems, so proving that something will always work is not usually the done thing.
 
It’s stupid but ok. The best way to effect change in a democracy is to actually vote out the people in office. Get new prosecutors in there, new mayors, new sheriffs. If that’s where the problem is.

If you want to improve policing destroying their assets isn’t going to do it.
Voting is way overrated. Sure people should organize to vote and there is plenty of work to be done there but opposition, protest politics are very important as evidenced by the Civil Rights movement itself as well as dozen of examples across the world.

In the end when it comes to voting most people don't really have the power to influence, only through voting, the actual choices on the ballot so even before you vote your choices are confined to an acceptable spectrum of choices.

Sometimes you need to send a message and in this case I want a generation of officers to know that in a world of smartphones and social media they have to be very careful about their use of force and can't just expect their superiors to cover their ass when they get a little overzealous and that any instance of abuse can snowball into this kind of chaos if there isn't accountability.
 
One thing I’ve seen multiple times is IQ scores from something like elementary school.
As somebody who actually experienced all that stuff where nobody knew what the hell to do with me academically as a kid, I'm pretty good at spotting the bullshit in those stories. My favorite is when they say something like their IQ being an odd number like 161.
 
Voting is way overrated. Sure people should organize to vote and there is plenty of work to be done there but opposition, protest politics are very important as evidenced by the Civil Rights movement itself as well as dozen of examples across the world.

In the end when it comes to voting most people don't really have the power to influence, only through voting, the actual choices on the ballot so even before you vote your choices are confined to an acceptable spectrum of choices.

Sometimes you need to send a message and in this case I want a generation of officers to know that in a world of smartphones and social media they have to be very careful about their use of force and can't just expect their superiors to cover their ass when they get a little overzealous and that any instance of abuse can snowball into this kind of chaos if there isn't accountability.
Yeah I’m not saying protests aren’t appropriate and helpful. You need to organize people to make them get out and vote for change. You need to make the voices heard.

But I think being arrested for murder is a much stronger deterrent than seeing cars you don’t own set on fire. Destroying property only hurts the overall cause of improving policing because it wastes resources; it doesn’t make any officers do anything except push back stronger.

Riots don’t send any message about use of force, they justify use of force. The arrests send the message.
 
seen on twitter....
"103k dead, pandemic still raging, cities across America burning, National Guard helicopters flying into cities to battle citizens, and #BunkerBoy Trump hiding with the lights out. This is pretty much how I always envisioned year four of a Trump presidency to be."

https://twitter.com/intent/like?tweet_id=1267490961627717634
 
@Andy Capp in a sense most machine learning methods are bayesian. You are trying to find the conditional estimate of some parameter, given a bunch of observations and beliefs about the problem. For example, let's take searching an area. Let's say there's an object that is in one of a few different bins, and for each bin you have a prior probability that the object is in it, and there's a probability that you will find it if you look in the bin, and then there's a probability that you'll look in that bin. You can pretty easily construct a bayesian rule for searching this space. This could be your state, action pair at each step, just so long as the markov property holds.
 
Status
Not open for further replies.
Back
Top