It mainly knew how to try for strikes, probably because it had way more opportunities to attempt them and the reward was greater. My guess is that the algorithm would figure out getting spares eventually. It'd just take a lot longer, both because of fewer opportunities to learn and because it would always be in competition with continued improvement in getting strikes. A perfect AI would never need to bowl for spares in the first place.
You have no idea how big of a problem that is. If it does one thing so perfectly that it never gets to experience anomalies. That's an inherent problem with fixed data inputs from simulated environments. That's why Tesla is using customer driving data to train their AI, the real-world data is extremely varied, potential inputs are limitless. They are teaching it the entire visual human experience and letting the NNs figure out how to navigate those experiences as safely as possible. It will be awesome to see the final NN in place, planning and control. Exciting.
It's simply not trained to handle incomplete pin sets. Since most of its throws are strikes, and they normally get rewarded more, it probably doesn't find any real value in training that specific rare scenario.
@@junkyyard2273 I'd say the core idea behind the strategy has more to do with spinning permitting one to hold the ball for longer, thus giving them more time to accelerate it. The centripetal force by itself does not contribute any energy, as it is orthogonal to the velocity.
@@junkyyard2273 it might be i remeber hearing centrifugal force isn’t really a thing in physics but I can’t remember the context of it I just know centripetal is the force keeping the object from flying away like a rope or this dudes arm in this case
My university occasionally has ice bowling events, literally bowling on an ice rink, wearing hockey gear, and this AI flopping around and faceplanting after throwing the ball is EXACTLY what it looks like 90% of the time
@@ihavenoson3384 "Hey how do we make ice's low friction fun" "How about using a ball that won't even be touching the floor" "but, that defeats the whole poi-" "ELASTIC BALL GO BRRRRRRRRRRRR"
As a bowler myself, this is hilarious. By the way, it might have been helpful to include information on which pins are standing as an input to the system. Then it might stand a chance at getting spares somewhat consistently. Bowling scoring is very strange, but basically it heavily punishes leaving an "open" frame, meaning you didn't get a strike or spare. This means that being able to pick up your spares is super important for good scores. Another thing to consider is that the hand can act like a lever, allowing you to release your thumb first, then use your fingers to add rotation to the ball. If the model had some sort of basic "hand" that essentially has hold of the ball in two locations, that might allow it to add a fair bit of rotation. Also, did you consider adding oil to the lane? A standard bowling lane will have oil applied to it, with most of it in the first 40 feet and biased toward the center of the lane. Then you could increase the friction between the ball and the surface to compensate.
You know... "Birthday bowling" was a big thing when I was a teenager. I never liked it much really, but here's the thing: Nobody EVER told me any of those things. Everybody was just supposed to know how bowling "works" , I am now a little bit wiser
@@tanelehala6422 Well here it did because the actual objective wasn't even part of the equation. Should really have started with "kock down these quills with the ball" rewarded and "don't cross that line with your feet" punished. In this case, the AI didn't settle for "good enough", the AI tried to maximize the amount of points it could get, and it was rewarded with points for things that don't grant you any in reality when bowling.
To be honest completely throwing yourself off balance is a pretty effective way to transfer energy to the ball. The only real difference is in real life people tend not to like to severely injure themselves on every frame or two.
i like how one of the initial objectives was finishing the episode standing upright so it wouldn't just instantly faceplant every time and then halfway through you went "actually you know what little buddy you just do what you want"
By the end is when the head being held high should have been started to be rewarded, the importanter shit was figured out from there just condition that model to also stay upright and you’re golden
I think a smarter modifier is that it should be punished when it lets it's head, thighs, and torso hit the floor. Though this should also be integrated into it's existing network so it knows how to bowl first.
The drastic strategy changes when there were only a few pins remaining is a testament to how well you tuned the reward parameters. If the reward for knocking down pins hadn't been tuned effectively, the AI never would've attempted spares, and instead would've just launched the ball down the center of the lane during a split to be rewarded for speed and direction. Well done!
Imagine seeing someone do this IRL and losing to them they just either get a strike or go in the other peoples lane or hits someone and no matter how good or bad the shot they just fall over
Heck, I only know 3 or 4 people that can put up a 130 or better every time and two of them were on a bowling team, I'd lose here 4/5 games for sure. Maybe losing to 'em wouldn't be as bad as sharing the lane with someone who needs an entire ER in the alley after every frame. . .
Oh my god, please do a video (or a series of videos!) where you train different AIs independently and have them compete against each other in a tournament structure. They can even continue to train a certain number of predetermined times after every game so we can see the best ones progress and gradually refine. I N-E-E-D to see this!
A 6 year old relative of mine used to bowl kind of like that. In order to get enough ball momentum he usually ended up falling down after release. It became part of his technique. His parents were used to his dare-devil energy in general such that they gave up trying to tame him. When you're 6, you can get away with that. (When he's 55 and his back aches, he might remember why.)
This makes me want an olympic bowler to perfect a technique that looks this dumb and win with it. Maybe keep it in their back pocket until they're ahead enough that they can take the risk.
Thank you so much for including all these little WiiSports references, especially the backwards-throw one at the end :D brought back memories from more than a decade ago with hours of laughter playing this game at a birthday party
This reminds me vividly of the time i went bowling with friends and i slipped while rolling, let go of the ball, and nearly faceplanted with watching the ball honest to god bounce two lanes over. Art imitates life.
I'm genuinely curious just how fast the ball could've been thrown had it not given up on aiming and kept increasing in speed, like I'm not expecting a rail gun but its funny to imagine them potentially throwing the ball with enough force to hit the pins directly without rolling.
@@tangentfox4677 It doesn't appear that the United States Bowling Congress, at least, has explicit rules which would cover doing so, though it seems likely to run afoul of rules regarding damaging the lane, pins, and other equipment.
@@coyoteseattle I suspect it's not in the rules because nobody thinks a human capable of it... For now. I eagerly await the screams of dismay from my local bowling alley the day they realize they were wrong.
@@karnewarrior apparently an olympic shot put weighs 16 lbs.. bowling balls typically weigh less than that, they're just bigger I guess- still seems pretty doable when you consider how far they're able to throw a shot put
@@karnewarrior You can absolutely chuck a 8 lb bowling ball down the lane without it touching the ground, but don't expect to not get kicked out after doing it.
The actual game at the end scared me. It uses the same exact technique I do, and I was sure to be the only one. No one else had the drive to send their entire body into the shot, let alone commit your face to it. AI is definitely taking over.
I love that the camera still zooms into the lane when they miss, it makes it so funny. It zooms in, and you expect to see the ball majestically appear and destroy those pins, but there’s just nothing.
The moment you mentioned rewarding higher speed with more reward, I just knew the guy was gonna throw that thing as hard as possible. I'm just impressed it didnt end up abusing the physics engine for that goal. Perhaps CodeBullet has spoiled me
If left for a long time it would have absolutely ended up abusing the physics engine somehow. But the more important point is correctly defining the rewards. Number of knocked down bowling pins should absolutely be a multiplier on the score. Zero pins = zero score. But even then knocking down half at superspeed might give more score that knocking down all at low speed, which is still stupid.
Thank you for these videos, for real man. I've been watching your stuff for months and they've indirectly inspired me to go to college for software engineering. I'll come back to this video in a few years once I finish up and let you know how it went.
The most hilarious thing about watching this is my irl group of friends in early adulthood would go bowling, and one of our friends in the group had professional training- and at one point, the way he would roll the ball would sometimes have him "fall over" afterwards. So another friend started intentionally falling over dramatically every time he rolled the ball to poke fun at him... and then it became a "meme" about him falling over and flopping like a fish. But the dude could out bowl any of us without effort...
The technique seems to be starting by pushing the ball outside into a T pose, half a circle downwards, and then loose forward. To compensate that inertia, one leg crosses behind. Interesting, I will try that.
That's not how engineering works! You keep trying until you run out of money, then ask for more. Sell exclusive merchandise to your fanbase for a quick boost of revenue. Launch one of your products into space for free direct promotion using your actual products. Have a plan and be prepared to change it, often.
A reminder that computers are stupid, and we expect a human brain to be able to overcome challenges an AI consistently fails at. Example: throwing a ball without committing ones face to the floor in holy matrimony.
You seem like an experienced programmer, visualisation developer, data analyst, video game designer, teacher, shitposter and engineer who also dabbles in drawing, cinematography, modelling, computer graphics, writing, user interface, publishing papers and gaming. Oh, and I also have this suspicion that you're an AI developer.
As someone who writes similar code for my games, I have to say that this video is brilliant. It's hilarious, but also does a good job of explaning the concepts of rewards, penalties and weights.
The 6th frame was actually pretty impressive, because the AI aimed for the exact right spot it needed to in order to make the spare. There just wasn’t enough speed and power to knock them down, but it is aiming and strategizing like a bowler.
Dunno,, I think the programing of the pins may have been off a bit or something. As you said, that ball was in the spot needed to make the spare. I typically aim to the other side of the front pin in that situation to get more direct pin movement backwards to get the sleeper. But the shot the AI took should have taken down more than the front pin.
I award an amazing 9 out of 10 to the AI for accuracy, but heck I have to award at least 10 out of 10 for its sheer commitment to the game on each and every turn.
The Bowler knows where his ball is is at all times. He knows this because he knows where it isn't. By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the ball from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
But a ball with spin on a complex axis (not just one of the X, Y, or Z axes) will follow a nonlinear path, as the spin axis itself is re-oriented by its friction with the floor. There would need to be some sort of polynomial algebra involved to solve for this differential, as there are a lot more than two variables (matrices, consisting of movement vector, gravity vector, spin, and floor-contact friction, in this case)
It's interesting seeing how it reacts to the pins. When it had a 7-10 split, it went right down the middle. But both times when it had only 1 pin up it tried to ride the right edge, even though one time the pin was on the left. I'm assuming it learned that sometimes it can pick up spares if it does that, but it's not aware of pin position so it's just going for what worked a different time. Which means that it will factor its strategy based on pins remaining, at least.
I was thinking more that the product of the rewards lead the doll to the behavior "aim straight, since spin is sufficient for deviations from a straight path". The doll is right-handed, so 10 pin? No problem. 7 pin? Spin it, but do not release the ball at an angle, because that virtually always leads to a bad outcome.
the whole 9 frames has look & feel of wii, from the semi-transparent bowler to the other players sitting, the visuals on strike/spare, the pin inset in the corner...
I think "definitely not forgetting" to include a reward system for knocking over the pins from the beginning should have warranted a fresh rewrite of the entire program. Having that in place from the first throw could have had an exponetial effect on the AI's ability to hit the pins much earlier on. Granted, it may have resulted in the AI throwing the bowling ball like a bullet, but who hasn't attempted that?
There are different league sanctions, but most require you to hold the ball a certain way and throw underhanded. You can even be disqualified for not putting your thumb in the hole. Of course, if you are open bowling, you can do anything you want short of violating the house rules.
@@nomisukeindustries As has been stated, most major bowling organizations don't restrict how you throw. If someone was extremely good at throwing overhand, they'd be allowed to do it.
I can absolutely vouch for the "Jazz Hands" technique. After many decades of bowling, I've learned that the "Jazz Hands" technique absolutely works like a charm. And judging from the way my friends bowl, the "flopping around the floor while flinging the ball up into the air in random directions" totally works too. Although as a spectator, you really have to learn to dodge behind the furniture and other obstacles pretty quick. Go bowling with my friends a few times, and you'll become a pro "dive behind stuff dodger". The driving range is another wonderful place where the jazz hands and diving out of the way techniques also work.
I’d use this for reward: pins knocked multiplied by head distance to ground, where 1.0 is slightly below regular height, 0.0 is slightly above the ground, and clamp it within 1 and 0
It's amazing how AI can really help improve and refine things in ways we might never have thought of before. I am looking forward to trying this technique the next time I go bowling! The other bowlers won't stand a chance.
What about a movie like Real Steel but instead of boxing it's robot bowling and they have to bring a geriatric bowler out of retirement to shadow bowl with his junkyard bot
Very nice. Also a reward for no contact with the ground other than the "foot". He might try and stay vertical. It was very interesting to see what bowling might look like if we didn't care about anything other than rolling a strike.
Suggestion: AI learns to play Yomi Hustle. Yomi Hustle is a turn-based fighting game where each player is given anywhere between a few seconds to a half hour to plan out the action of their fighter down to the frame-data. The goal is to make AI that master Yomi hustle, analyze the game balance (overall win/loss ratio against all characters and best/worst win/loss ratio between them), then select a character mod and run it through to see if it fits that balance.
Good Idea, but 1. b2 doesnt own the rights to YOMIH 2. YOMIH is very complicated, even compared to stuff like this, as it has many, many inputs and outputs that would need to be considered 3. b2 would have to program his own version of YOMIH in order to give the AI precise inputs such as the velocity and position of the players, and it would be very difficult to make a good recreation as b2 wouldnt know the numbers to throw in for YOMIHs engine 4. The computing power required for this would also be insane, as the AI would have to predict every single scenario and weigh the likelyhood of whats gonna happen, and somehow decide as to which move would be the best, which would also require looking even more moves in the future in order to correctly see if the enemy would be able to retaliate easily, or if the AI would be able to combo them. YOMIH does seem like a good game to train an AI for, but even after hundreds of hours being put into making it, i doubt the AI would even be able to pull off basic combos, and thats not very good material for a video.
Toribash seems like a good candidate: * f2p for the past 16 years, * Input interface is 20 joints with 4 possible states each. Not a small space, but simple to handle and discreet, * Observation space is continuous - with position, angle (and speed) of all body parts, * Leans into the theme of this and the previous video - 3d ragdoll physics
@@Snowowowie 1. Didn't realize that was a requirement. Why didn't that apply for all the other established games he's done videos for? Monopoly in the creative commons? 2. I have seen AI take on the likes Starcraft and Pokémon, complication is only a matter of his system's capacity. 3. I know that is not a strict requirement by virtue of the other games he and others have tested. It is a requirement for lesser systems, but it's far from an absolute. Even if it was though the good thing about it being a turn-based fighting game is that most of the values are evident up-front, and we've already seen him crack into games like Rocket League for specifics. 4. That is the same problem for every turn-based game from as early as chess. The goal is not to plan out every move, it is to select the most optimal within a range for the immediate situation. As for how it judges that, that's were the machine "learning" comes in, along with some basic conditions to define value. Throw moves at the wall and see what sticks. As for the computing to understand the moves, Yomi Hustle already has a built-in system to literally let you check every single move down to its frame data against every single one of the opponent's, might ease up on the burden. Yomi Hustle is only complicated in a vacuum. As long as it has the capacity comparable to that which learns the most traditional of board games, after hundreds of hours it would probably look just like them: practically impossible for humans to beat without handicaps.
The one extra thing I'd liked to have seen would be a stability reward for how much time the AI was able to spend standing after releasing the ball (scaling with the number of pins hit, so as not to create a conflicting incentive). Adding it wouldn't invalidate any of the work that had come before and, as funny as it is to watch it faceplant every time, it would be cool to eventually see it master all aspects of the simulation, especially with those terrible feet.
a few pointers: keep it simple. reduce your reward function to just knocking over the pins + staying in the right lane. clamp lane reward to 1.0 (or low exp towards center) until sides, then go negative exp or fixed to a value once over bounds. if you want speed, give a reward reverse proportional to time spent until pin hit. all rewards must be lower in total than pin knockdown reward. that should avoid local minima, maximize knocking pins down + keep optimizing for speed after pins are knocked down. don't reward means, only outcome. if spin is benficial it will learn it, if pin position is benficial it will learn it as well.
A couple other interesting parameters to add might be to simulate pain/injury avoidance by punishing it for falling over, and also reward points not just for being in the right lane but only if the ball is touching the lane. Maybe that would discourage falling over and discourage throwing the ball through the air out onto the lane. Maybe that would result in more familiar bowling technique, although it would be more boring too.
Honestly, the AI getting a spare is more impressive than a strike.
uh huh
I was willing that ball on
It mainly knew how to try for strikes, probably because it had way more opportunities to attempt them and the reward was greater. My guess is that the algorithm would figure out getting spares eventually. It'd just take a lot longer, both because of fewer opportunities to learn and because it would always be in competition with continued improvement in getting strikes. A perfect AI would never need to bowl for spares in the first place.
You have no idea how big of a problem that is. If it does one thing so perfectly that it never gets to experience anomalies. That's an inherent problem with fixed data inputs from simulated environments. That's why Tesla is using customer driving data to train their AI, the real-world data is extremely varied, potential inputs are limitless. They are teaching it the entire visual human experience and letting the NNs figure out how to navigate those experiences as safely as possible. It will be awesome to see the final NN in place, planning and control. Exciting.
It's simply not trained to handle incomplete pin sets. Since most of its throws are strikes, and they normally get rewarded more, it probably doesn't find any real value in training that specific rare scenario.
I was hoping to see the AI figure out centrifugal force and start spinning faster and faster each time to launch the bowling ball at Mach speed
Avoid having the ball hit the lane altogether, just YEET the ball at the pins
isnt that just centripetal force at least in this case?
@@junkyyard2273 I'd say the core idea behind the strategy has more to do with spinning permitting one to hold the ball for longer, thus giving them more time to accelerate it. The centripetal force by itself does not contribute any energy, as it is orthogonal to the velocity.
@@junkyyard2273yes, but meh
@@junkyyard2273 it might be i remeber hearing centrifugal force isn’t really a thing in physics but I can’t remember the context of it I just know centripetal is the force keeping the object from flying away like a rope or this dudes arm in this case
The Wii Bowling imitation was immaculate. This entire project is immaculate. I can't wait to see more from you!
You are gonna love the next video
@@b2stud I bet!
The ending is just spot on!
E
extreme demon
4:45 as a bowling prof i can confirm that this technique is flawless
LOL
I read this as "bowling professor," and I think that's a solid title for you, sir.
@guccilibrarian4728 is it not bowling prof?
There's a professor for bowling? 🧐
@@onlinefriend3889 probably
I'm going to have to try that "throw ball and immediately faceplant on the floor" technique next time I go bowling.
Record it please
you HAVE to upload that PLEASE
Don't forget to use the elasticity of your spine.
Don't forget to add the little swimming on asphalt move
It seems like a perfectly fine technique for a single throw, it just impairs your performance for subsequent throws.
My university occasionally has ice bowling events, literally bowling on an ice rink, wearing hockey gear, and this AI flopping around and faceplanting after throwing the ball is EXACTLY what it looks like 90% of the time
Is there anywhere I can watch this? Sounds hilarious
@@b2stud The university's recreation department (UVM Campus Rec) has a video of it up here on CZcams, idk where else
I do such on ice even when not trying to bowl.
@@b2stud czcams.com/video/3xW0ue2w3lI/video.html The ball they use is elastic as hell.
@@ihavenoson3384 "Hey how do we make ice's low friction fun"
"How about using a ball that won't even be touching the floor"
"but, that defeats the whole poi-"
"ELASTIC BALL GO BRRRRRRRRRRRR"
I love how the AI resorts to faceplanting into the ground and throwing a tantrum every time it throws the ball 😂😂
It actually reminds me of an old gag - the Machuga flop (PLEASE Google that)
That's part of the game, surely?!
As a bowler myself, this is hilarious.
By the way, it might have been helpful to include information on which pins are standing as an input to the system. Then it might stand a chance at getting spares somewhat consistently. Bowling scoring is very strange, but basically it heavily punishes leaving an "open" frame, meaning you didn't get a strike or spare. This means that being able to pick up your spares is super important for good scores.
Another thing to consider is that the hand can act like a lever, allowing you to release your thumb first, then use your fingers to add rotation to the ball. If the model had some sort of basic "hand" that essentially has hold of the ball in two locations, that might allow it to add a fair bit of rotation.
Also, did you consider adding oil to the lane? A standard bowling lane will have oil applied to it, with most of it in the first 40 feet and biased toward the center of the lane. Then you could increase the friction between the ball and the surface to compensate.
You know... "Birthday bowling" was a big thing when I was a teenager. I never liked it much really, but here's the thing:
Nobody EVER told me any of those things. Everybody was just supposed to know how bowling "works" , I am now a little bit wiser
I love how he finds a technique that works perfectly.. and then throws it into the next lane a second later lmao
That's what makes AI better, it never stops searching for better results, while humans will settle for the best result their mind judges it to be so
@@colinhaml but as said in this video AI does sometimes settle for "good enough"
@@tanelehala6422that's the age old problem of exploration vs exploitation
@@tanelehala6422 Well here it did because the actual objective wasn't even part of the equation. Should really have started with "kock down these quills with the ball" rewarded and "don't cross that line with your feet" punished. In this case, the AI didn't settle for "good enough", the AI tried to maximize the amount of points it could get, and it was rewarded with points for things that don't grant you any in reality when bowling.
The Jeopardy game show AI (a decade ago) did something similar. After an opponent answered incorrectly, the AI then gave the same incorrect answer.
To be honest completely throwing yourself off balance is a pretty effective way to transfer energy to the ball. The only real difference is in real life people tend not to like to severely injure themselves on every frame or two.
Sounds like you are not a serious bowler, the pros use this move on occasion
Not if you want any amount of accuracy.
E
Cowards
Speak for yourself
That last throw in the 10th frame sent me into hysterics. AI just hauls off, full send cranks it to the right...only for it to hop over to the left.🤣
he said "nope, that lane's no good!"
For all we know, if stumps for feet is all we had, this could be the best technique to bowl.
i like how one of the initial objectives was finishing the episode standing upright so it wouldn't just instantly faceplant every time and then halfway through you went "actually you know what little buddy you just do what you want"
By the end is when the head being held high should have been started to be rewarded, the importanter shit was figured out from there just condition that model to also stay upright and you’re golden
My mans was created with spherical feet too so I saw it as a fruitless endeavor lol
Yea lol
@@Enter54623 least contact with floor as possible. watch him learn to jump shot a bowling bowl ))
I think a smarter modifier is that it should be punished when it lets it's head, thighs, and torso hit the floor.
Though this should also be integrated into it's existing network so it knows how to bowl first.
I love how the camera still zooms in on the pins even though the bowling ball wasn't even remotely close to entering the correct bowling lane
Same, also I like your name.
are you itchy
@@gabrielharrison9800 why thank you
I read this JUST as it happened lmao
@@PineappleElephant68 yes
I love how it starts in such an organized pose and then immediately throws itself into the ground
The drastic strategy changes when there were only a few pins remaining is a testament to how well you tuned the reward parameters. If the reward for knocking down pins hadn't been tuned effectively, the AI never would've attempted spares, and instead would've just launched the ball down the center of the lane during a split to be rewarded for speed and direction. Well done!
Imagine seeing someone do this IRL and losing to them
they just either get a strike or go in the other peoples lane or hits someone
and no matter how good or bad the shot they just fall over
"Yeah, I tried to learn to bowl standing up, but I could never let go of the ball."
Heck, I only know 3 or 4 people that can put up a 130 or better every time and two of them were on a bowling team, I'd lose here 4/5 games for sure. Maybe losing to 'em wouldn't be as bad as sharing the lane with someone who needs an entire ER in the alley after every frame. . .
It is night, I must be quiet. These comments are fucking killing me.
Sure, they might win the game, but then they have to pay their hospital bills.
no one would question this behavior
Bowling Alleys are basically bar substitutes. 😆
I love how he is so exhausted after every throw that he just straight up dies
Pretty much me at the bowling alley.
I sympathise. If you don't get any points for standing up, why bother?
I'd like to see real people bowl with spheres for feet
He just avoids the cameraman at all costs!
yeah seems like he should have rewarded staying upright after throwing more
Oh my god, please do a video (or a series of videos!) where you train different AIs independently and have them compete against each other in a tournament structure. They can even continue to train a certain number of predetermined times after every game so we can see the best ones progress and gradually refine. I N-E-E-D to see this!
Lol. This AI went 4 point over my PR. Love these AI vids. Have an upvote
You too
Have you tried falling over
I honestly think the most impressive part was him picking up the spare in frame 5
especially since when it was the 7 pin left he goes wild and chucks the ball into the wrong lane
Had me feeling for the lil bro. Like way to go bud!
I don't want to add an extra like to this, cause you're sitting at 300!!!
@@daveg5857 Nobody cares. Comment likes are not achievements.
@@CooManTunes I said that because 300 is a perfect game in bowling.
A ragdoll that barely knows the concept of sports, balance, or even breathing is somehow better than me at bowling.
That's because those concepts are holding you back
To be fair people who don't play bowling are like the first neural net. Source: Myself.
it might be better than you at bowling, but you are probably better at not flopping around like a fish on the ground after every throw.
@@BrowncoatFairy overrated skill tbh
@@ARVash It's amazing how self-existence (not even preservation) is such a limiting factor to so many amazing skills.
Therapist - "you can't hear words"
Wii Sports Bowling - Strike!
"NIcE tHrOw🤓"
Someone needs to make a live stream of AI learning, this is so soothing and fun to watch
"I forgot how to walk"
*having the most painful seizure*
Really opened my eyes when it comes to bowling. All you need is a lack of self preservation and you can do anything!
oooooh so that's why i'm so good at bowling
"the man who is willing to commit suicide has the initiative" is a common phrase in Russian bowling leagues
A 6 year old relative of mine used to bowl kind of like that. In order to get enough ball momentum he usually ended up falling down after release. It became part of his technique. His parents were used to his dare-devil energy in general such that they gave up trying to tame him. When you're 6, you can get away with that. (When he's 55 and his back aches, he might remember why.)
This makes me want an olympic bowler to perfect a technique that looks this dumb and win with it.
Maybe keep it in their back pocket until they're ahead enough that they can take the risk.
@@macaronandcheese1811 LMAO!
Thank you so much for including all these little WiiSports references, especially the backwards-throw one at the end :D
brought back memories from more than a decade ago with hours of laughter playing this game at a birthday party
This reminds me vividly of the time i went bowling with friends and i slipped while rolling, let go of the ball, and nearly faceplanted with watching the ball honest to god bounce two lanes over.
Art imitates life.
I'm genuinely curious just how fast the ball could've been thrown had it not given up on aiming and kept increasing in speed, like I'm not expecting a rail gun but its funny to imagine them potentially throwing the ball with enough force to hit the pins directly without rolling.
Is.. is that legal?
@@tangentfox4677 It doesn't appear that the United States Bowling Congress, at least, has explicit rules which would cover doing so, though it seems likely to run afoul of rules regarding damaging the lane, pins, and other equipment.
@@coyoteseattle I suspect it's not in the rules because nobody thinks a human capable of it... For now.
I eagerly await the screams of dismay from my local bowling alley the day they realize they were wrong.
@@karnewarrior apparently an olympic shot put weighs 16 lbs.. bowling balls typically weigh less than that, they're just bigger I guess- still seems pretty doable when you consider how far they're able to throw a shot put
@@karnewarrior You can absolutely chuck a 8 lb bowling ball down the lane without it touching the ground, but don't expect to not get kicked out after doing it.
The actual game at the end scared me. It uses the same exact technique I do, and I was sure to be the only one. No one else had the drive to send their entire body into the shot, let alone commit your face to it. AI is definitely taking over.
yeah. Its mostly about you.
they should allow this as an option to play against at an actual bowling alley. it'll fit right in with the bowling animations
I cannot overstate how much I love that the AI proceeds to just fucking die every time after it throws the ball.
That 10th frame was an emotional roller coaster.
Imagine if anyone makes a 0-10-0 on the 10th.
Always keep them guessing
It's like the nerves really got to him
I personally enjoyed the spare pickup on the 8th.. just "nope f this roll".
I love that the camera still zooms into the lane when they miss, it makes it so funny. It zooms in, and you expect to see the ball majestically appear and destroy those pins, but there’s just nothing.
Wii sports also does it, which the final result is entirely mimicking (for anyone who hadn't figured that out)
It's like it's saying "bruh..."
@@FairyNuffMuffin2 I loved how he included the backward-roll and the crowd yell at the end. Nostalgia hahaa
9:45 HAHAHA how he tried to catch himself after the fling.
The way the AI chose to throw the ball had my crying I was laughing so hard!!😂😂
I appreciate the presense of little ball spectators. Makes the environment look way more lively rather than looking like setting for the experiments.
Not gonna lie, I got way too distracted by the $33 jellybean on sale at the bar.
☠️☠️☠️☠️☠️☠️☠️
Can you imagine being a ball person and just watching this dude spazz out in the bowling alley for like a week?
E
@@Jokoko2828 :
*I would be more upset the AI was throwing my buddy around.*
The moment you mentioned rewarding higher speed with more reward, I just knew the guy was gonna throw that thing as hard as possible. I'm just impressed it didnt end up abusing the physics engine for that goal. Perhaps CodeBullet has spoiled me
If left for a long time it would have absolutely ended up abusing the physics engine somehow.
But the more important point is correctly defining the rewards. Number of knocked down bowling pins should absolutely be a multiplier on the score. Zero pins = zero score. But even then knocking down half at superspeed might give more score that knocking down all at low speed, which is still stupid.
*Clips ball into chest*
*Launches ball directly into pins at mach 6*
Me when I play games.
Would have had to program a trip to the gym between frames.
Was waiting for this comment lol
Thank you for these videos, for real man. I've been watching your stuff for months and they've indirectly inspired me to go to college for software engineering. I'll come back to this video in a few years once I finish up and let you know how it went.
3:25 This isn't AI. This is just raw footage of me bowling. I'm in this video and I don't like it.
10:09 the ai synchronized hitting the floor with the sound of the ball slamming down got me XD
10:50 is even better he just slams his face into the ground the moment the ball lands
The most hilarious thing about watching this is my irl group of friends in early adulthood would go bowling, and one of our friends in the group had professional training- and at one point, the way he would roll the ball would sometimes have him "fall over" afterwards. So another friend started intentionally falling over dramatically every time he rolled the ball to poke fun at him... and then it became a "meme" about him falling over and flopping like a fish. But the dude could out bowl any of us without effort...
Sorry to break the news to you like this... Your friend is an AI...
Your friends group sounds like a riot 😂
liar
Imagine having your claim to fame that you could out-bowl your friends, but have to fall over and embarrass yourself in order to do it.
@@B3Bandwould you rather...?
The technique seems to be starting by pushing the ball outside into a T pose, half a circle downwards, and then loose forward. To compensate that inertia, one leg crosses behind. Interesting, I will try that.
The edits! Insane editing, genuinely so clean
I always hate it when I go to a bowling alley and realise I forgot how to do anything, so I have to have my brain rebuilt from scratch.
Average Tuesday night for me
2038 moment
Using AWD after you drunk too much... know these days...
It’s nice to know that it’s perfectly natural to just go “meh good enough, I tried”
That's not how engineering works! You keep trying until you run out of money, then ask for more. Sell exclusive merchandise to your fanbase for a quick boost of revenue. Launch one of your products into space for free direct promotion using your actual products. Have a plan and be prepared to change it, often.
A reminder that computers are stupid, and we expect a human brain to be able to overcome challenges an AI consistently fails at. Example: throwing a ball without committing ones face to the floor in holy matrimony.
im only halfway through this video and im thoroughly entertained between in-progress training ai being goofy and your hilarious commentary.
Wondering if anyone noticed that a single jellybean is $33
You seem like an experienced programmer, visualisation developer, data analyst, video game designer, teacher, shitposter and engineer who also dabbles in drawing, cinematography, modelling, computer graphics, writing, user interface, publishing papers and gaming. Oh, and I also have this suspicion that you're an AI developer.
pretty accurate lol
Best we can offer is $17/hr
The only reason to pay that much is if he had fifteen years of experience. Give him the entry rate of 16.50.
@@wangsunfuh8889 I think youtube+patreon pay out a lot less than that, unfortunately.
@@dakodahuber 17/h sounds a lot man, where do you find these clients? 😛
You should do a 24/7 stream of AI bowling, or all sorts of training.I know I’d watch it from time to time
Would watch
Might I suggest the AI vtuber that is constantly learning: Neuro-sama!
@@comicsans2516 she doesnt do bowling tho
@@nickku_ Just tell her creator to get on it. He already did it with among us
@@dean_l33 Vedal would rather spend the time making her troll better
As someone who writes similar code for my games, I have to say that this video is brilliant. It's hilarious, but also does a good job of explaning the concepts of rewards, penalties and weights.
I know the math is cool and all but the resulting animations of flop-bowling are really the highlight of this video
That face plant after every shot kept bringing me joy. I really thought it would get old, but it just never did.
The 6th frame was actually pretty impressive, because the AI aimed for the exact right spot it needed to in order to make the spare. There just wasn’t enough speed and power to knock them down, but it is aiming and strategizing like a bowler.
Dunno,, I think the programing of the pins may have been off a bit or something. As you said, that ball was in the spot needed to make the spare. I typically aim to the other side of the front pin in that situation to get more direct pin movement backwards to get the sleeper. But the shot the AI took should have taken down more than the front pin.
I award an amazing 9 out of 10 to the AI for accuracy, but heck I have to award at least 10 out of 10 for its sheer commitment to the game on each and every turn.
I love watching ai learn and trying new stuff to get things right.
The Bowler knows where his ball is is at all times. He knows this because he knows where it isn't. By subtracting where it is from where it isn't, or where it isn't from where it is (whichever is greater), it obtains a difference, or deviation. The guidance subsystem uses deviations to generate corrective commands to drive the ball from a position where it is to a position where it isn't, and arriving at a position where it wasn't, it now is.
Wait istg I've heard smth like this before. What's it a reference to?
@@crusatyr1452 missile kalman filter video?
But a ball with spin on a complex axis (not just one of the X, Y, or Z axes) will follow a nonlinear path, as the spin axis itself is re-oriented by its friction with the floor. There would need to be some sort of polynomial algebra involved to solve for this differential, as there are a lot more than two variables (matrices, consisting of movement vector, gravity vector, spin, and floor-contact friction, in this case)
@@revimfadli4666 idk what that is ._.
@@crusatyr1452 Missile Guidance for Dummies
It's interesting seeing how it reacts to the pins. When it had a 7-10 split, it went right down the middle. But both times when it had only 1 pin up it tried to ride the right edge, even though one time the pin was on the left. I'm assuming it learned that sometimes it can pick up spares if it does that, but it's not aware of pin position so it's just going for what worked a different time. Which means that it will factor its strategy based on pins remaining, at least.
I was thinking more that the product of the rewards lead the doll to the behavior "aim straight, since spin is sufficient for deviations from a straight path". The doll is right-handed, so 10 pin? No problem. 7 pin? Spin it, but do not release the ball at an angle, because that virtually always leads to a bad outcome.
I love how he looks so composed but then has a seizure immediately every time 😂
THANK YOU for including what's at the end!!!
Love the little Wii Sports jingle at 8:02.
the whole 9 frames has look & feel of wii, from the semi-transparent bowler to the other players sitting, the visuals on strike/spare, the pin inset in the corner...
10 frames duh
the song that plays after, as well as the ui, is all also a reference
I think "definitely not forgetting" to include a reward system for knocking over the pins from the beginning should have warranted a fresh rewrite of the entire program. Having that in place from the first throw could have had an exponetial effect on the AI's ability to hit the pins much earlier on. Granted, it may have resulted in the AI throwing the bowling ball like a bullet, but who hasn't attempted that?
It is perfectly legal for a bowler to throw the ball directly at the pins. It could even be thrown overhand.
There are different league sanctions, but most require you to hold the ball a certain way and throw underhanded. You can even be disqualified for not putting your thumb in the hole. Of course, if you are open bowling, you can do anything you want short of violating the house rules.
@@nomisukeindustries Neither USBC, BTBA, nor World Bowling make any such restriction.
@@nomisukeindustries As has been stated, most major bowling organizations don't restrict how you throw. If someone was extremely good at throwing overhand, they'd be allowed to do it.
@@merp9610I would pay top dollar to see this
The last few minutes had me laughing non-stop, had to catch my breath! The AI falling down, throwing silly shots, hilarious!
this video is absolute gold. so glad I found it
I can absolutely vouch for the "Jazz Hands" technique. After many decades of bowling, I've learned that the "Jazz Hands" technique absolutely works like a charm. And judging from the way my friends bowl, the "flopping around the floor while flinging the ball up into the air in random directions" totally works too. Although as a spectator, you really have to learn to dodge behind the furniture and other obstacles pretty quick. Go bowling with my friends a few times, and you'll become a pro "dive behind stuff dodger". The driving range is another wonderful place where the jazz hands and diving out of the way techniques also work.
for some reason this reminds me of rtgames john golf bit
This is unbeatable content! It's fun, it's visually appealing and the AI work behind is really complex. How the hell did you get so good?
Mate, just spent a lot of time alone
@@b2studAI bowling tournament?
@b2studios If that's all it takes then I should be a professional by now.
@@b2stud then why am I not good at ai already?
@@b2stud unsupervised learning ...
The "Shhhhwt" on toss is really the cherry on top.
You deserve a ton more viewers! What a cool project!
I will wait here patiently for 40 hours
Same...
Still waiting
Me too
I missed the premier
did you practice during that time?
Man I was really rooting for him on the 8th frame. I laughed so hard when he chucked the ball sideways! 😂
One of the funniest videos ever made. And interesting training sequence montages.
I love how the AI shits itself whenever there's only one pin remaining.
This is what true bowling mastery looks like. If you don't like it then you just aren't strong enough to handle these advanced techniques.
The throw at 9:09 is 100% me when I choke and cant take the pressure of downing a single, lonely pin for a spare
Lol that is so relatable
I’d use this for reward: pins knocked multiplied by head distance to ground, where 1.0 is slightly below regular height, 0.0 is slightly above the ground, and clamp it within 1 and 0
Just referred this video to a friend learning about Transfer Learning. The open brain surgery is the best explanation on it I have heard in a while.
It's amazing how AI can really help improve and refine things in ways we might never have thought of before. I am looking forward to trying this technique the next time I go bowling! The other bowlers won't stand a chance.
You’ll be floored with the results.
Coincidentally, you also won't be able to stand afterwards.
What about a movie like Real Steel but instead of boxing it's robot bowling
and they have to bring a geriatric bowler out of retirement to shadow bowl with his junkyard bot
9:28 literally had me dying- i don't know why the ball slowly creeping up on the pin was so funny to me, but i choked on my sprite laughing
comedic slowness
I was just impressed it actually went all the way to the pin
A sprite is a 2D game object
I think the pathetic thunk of the single pin being knocked over helps
@@The-EJ-Factor 🤓
They were talking about the drink.. 🙄
Very nice. Also a reward for no contact with the ground other than the "foot". He might try and stay vertical. It was very interesting to see what bowling might look like if we didn't care about anything other than rolling a strike.
my man really just fell down, he didnt feel anything, he dropped his ball then felt the pain of stepping in lego but its your whole body
Suggestion: AI learns to play Yomi Hustle. Yomi Hustle is a turn-based fighting game where each player is given anywhere between a few seconds to a half hour to plan out the action of their fighter down to the frame-data. The goal is to make AI that master Yomi hustle, analyze the game balance (overall win/loss ratio against all characters and best/worst win/loss ratio between them), then select a character mod and run it through to see if it fits that balance.
Good Idea, but
1. b2 doesnt own the rights to YOMIH
2. YOMIH is very complicated, even compared to stuff like this, as it has many, many inputs and outputs that would need to be considered
3. b2 would have to program his own version of YOMIH in order to give the AI precise inputs such as the velocity and position of the players, and it would be very difficult to make a good recreation as b2 wouldnt know the numbers to throw in for YOMIHs engine
4. The computing power required for this would also be insane, as the AI would have to predict every single scenario and weigh the likelyhood of whats gonna happen, and somehow decide as to which move would be the best, which would also require looking even more moves in the future in order to correctly see if the enemy would be able to retaliate easily, or if the AI would be able to combo them.
YOMIH does seem like a good game to train an AI for, but even after hundreds of hours being put into making it, i doubt the AI would even be able to pull off basic combos, and thats not very good material for a video.
Toribash seems like a good candidate:
* f2p for the past 16 years,
* Input interface is 20 joints with 4 possible states each. Not a small space, but simple to handle and discreet,
* Observation space is continuous - with position, angle (and speed) of all body parts,
* Leans into the theme of this and the previous video - 3d ragdoll physics
And the ideal end is to be able to pit two of these bots against each other and have them play real-time matches.
@@Snowowowie 1. Didn't realize that was a requirement. Why didn't that apply for all the other established games he's done videos for? Monopoly in the creative commons?
2. I have seen AI take on the likes Starcraft and Pokémon, complication is only a matter of his system's capacity.
3. I know that is not a strict requirement by virtue of the other games he and others have tested. It is a requirement for lesser systems, but it's far from an absolute. Even if it was though the good thing about it being a turn-based fighting game is that most of the values are evident up-front, and we've already seen him crack into games like Rocket League for specifics.
4. That is the same problem for every turn-based game from as early as chess. The goal is not to plan out every move, it is to select the most optimal within a range for the immediate situation. As for how it judges that, that's were the machine "learning" comes in, along with some basic conditions to define value. Throw moves at the wall and see what sticks. As for the computing to understand the moves, Yomi Hustle already has a built-in system to literally let you check every single move down to its frame data against every single one of the opponent's, might ease up on the burden.
Yomi Hustle is only complicated in a vacuum. As long as it has the capacity comparable to that which learns the most traditional of board games, after hundreds of hours it would probably look just like them: practically impossible for humans to beat without handicaps.
@@Snowowowie I mean, unity remakes seem feasible to him
3:19 the tv in a bowling alley when you hit a strike
The craziest thing is, we watched this guy teach this thing in hours. In years...Bowling will be over with for all.
i have no idea who this guy is but this is one of the best youtube videos i have ever seen
From AI learns to swing like Spiderman to inventing bowling techniques
Next up: AI learns to take over the world
This is number 22
So Plague Inc. then?
Also turn that into 2 AIs and make them mice.
Ok pinky
@@youtubeuniversity3638
Or, god forbid, Grand Strategy games.
I just want you to know that I was quite literally crying from laughter during this. Thank you so much for sharing this experiment!
The one extra thing I'd liked to have seen would be a stability reward for how much time the AI was able to spend standing after releasing the ball (scaling with the number of pins hit, so as not to create a conflicting incentive). Adding it wouldn't invalidate any of the work that had come before and, as funny as it is to watch it faceplant every time, it would be cool to eventually see it master all aspects of the simulation, especially with those terrible feet.
Legend has it, the AI never stopped falling.
3:43 when he falls down then goes back up it reminds me of the big floppy guys you see in car dealerships
a few pointers: keep it simple. reduce your reward function to just knocking over the pins + staying in the right lane. clamp lane reward to 1.0 (or low exp towards center) until sides, then go negative exp or fixed to a value once over bounds. if you want speed, give a reward reverse proportional to time spent until pin hit. all rewards must be lower in total than pin knockdown reward.
that should avoid local minima, maximize knocking pins down + keep optimizing for speed after pins are knocked down. don't reward means, only outcome. if spin is benficial it will learn it, if pin position is benficial it will learn it as well.
I was a little surprised that he didn't prioritize pins down, myself.
i love watching early AI try to figure out what is going on. the back shimmy 5 pointer has to be a highlight moment
A couple other interesting parameters to add might be to simulate pain/injury avoidance by punishing it for falling over, and also reward points not just for being in the right lane but only if the ball is touching the lane. Maybe that would discourage falling over and discourage throwing the ball through the air out onto the lane. Maybe that would result in more familiar bowling technique, although it would be more boring too.
3:38 those pins falling are all in the leg twitch baby
Poor guy really felt the pressure in the 8th frame there… (@10:28)
It looked like a CGI version of me bowling 🤣
Falling flat on the face just to get a strike is the definition of "no pain no gain" 💀💀💀
"Yoo AI you got a strike!"
"No, I think i got a stroke"
Watching them faceplant every bowl made me laugh so hard i couldnt breathe through almost the whole video 😂 ☠️
It's too funny! One too many drinks for this guy.
Same! I haven't laughed that hard in months!! 😂
I'd be exhaused too if I were an AI under his command.
10:36 I love how the ai subtitle here just says "thank you"
Just thanking you for watching till the end
"OVER THE LINE!!! Mark it Zero. Sorry, Smokey. Your foot crossed the line."
Really impressive! Especially the frame 5 spare!