With ChatGPT (the latest open source AI educational app) causing rumblings in educational circles you have to wonder how long until AI enters sport.

My guess is that it is probably already used to some degree in NFL to predict the success probability of particular plays.

Could it have application in Australian rules to predict trends in say centre bounces. I don't know if AI can interpret video yet but imagine if it could.

For those who follow these things, humans lost the ability to beat computers at chess in around 2006 and even now you can download programs to your phone that will beat any human.

2 minutes ago, Diamond_Jim said:

With ChatGPT (the latest open source AI educational app) causing rumblings in educational circles you have to wonder how long until AI enters sport.

My guess is that it is probably already used to some degree in NFL to predict the success probability of particular plays.

Could it have application in Australian rules to predict trends in say centre bounces. I don't know if AI can interpret video yet but imagine if it could.

For those who follow these things, humans lost the ability to beat computers at chess in around 2006 and even now you can download programs to your phone that will beat any human.

I'm still winning @ 80% against my Backgammon app 😤 

https://www.the42.ie/western-bulldogs-machine-learning-3213937-Feb2017/
 

About 6 years ago 

25 minutes ago, DeeSpencer said:

https://www.the42.ie/western-bulldogs-machine-learning-3213937-Feb2017/
 

About 6 years ago 

Interesting... mainly used to measure their own players which makes sense as they could access the data.

Its use in recruitment is fascinating

32 minutes ago, DeeSpencer said:

https://www.the42.ie/western-bulldogs-machine-learning-3213937-Feb2017/
 

About 6 years ago 

maybe champion data should take up ai to improve their analysis. shouldn't be hard to better current efforts 😄

Mark Jackson had that.

AI will really kick in as a genuinely useful thing when it begins to be used to support 'mutual best outcome' estimation in things like trade week.

Not so much that it is cleverer or philosophically deep, but an AI would have the capacity to look at all clubs objectives simultaneously without getting a ripping headache, and without personally caring which club looked like they were 'doing best'.

Actually, that second part is an aspect of why Melbourne have pick-traded so effectively in the last few years - caring a lot less about who 'won' and just being interested in 'did we gain?'

Okay, I'll come clean - there's a thing my older brother (Yes, that would be, 'Big Goffy') was involved in and it is freak'n awesome. You'll need your nerd hat on:

https://ai.facebook.com/blog/cicero-ai-negotiates-persuades-and-cooperates-with-people/?utm_source=twitter&utm_medium=organic_social&utm_campaign=cicero&utm_content=video

 

We already have the resources in place, if you count Al Richardson.

1 hour ago, Little Goffy said:

AI will really kick in as a genuinely useful thing when it begins to be used to support 'mutual best outcome' estimation in things like trade week.

Not so much that it is cleverer or philosophically deep, but an AI would have the capacity to look at all clubs objectives simultaneously without getting a ripping headache, and without personally caring which club looked like they were 'doing best'.

Actually, that second part is an aspect of why Melbourne have pick-traded so effectively in the last few years - caring a lot less about who 'won' and just being interested in 'did we gain?'

Okay, I'll come clean - there's a thing my older brother (Yes, that would be, 'Big Goffy') was involved in and it is freak'n awesome. You'll need your nerd hat on:

https://ai.facebook.com/blog/cicero-ai-negotiates-persuades-and-cooperates-with-people/?utm_source=twitter&utm_medium=organic_social&utm_campaign=cicero&utm_content=video

 

Thanks for sharing that

found this comparison between ChatGPT and Claude interesting

Im dabbling in this space in health a little bit

 

Pffftt.. stupid AI, it did my last uni essay. 

Edited January 16, 20233 yr by Ethan Tremblay

1 hour ago, Little Goffy said:

AI will really kick in as a genuinely useful thing when it begins to be used to support 'mutual best outcome' estimation in things like trade week.

Not so much that it is cleverer or philosophically deep, but an AI would have the capacity to look at all clubs objectives simultaneously without getting a ripping headache, and without personally caring which club looked like they were 'doing best'.

Actually, that second part is an aspect of why Melbourne have pick-traded so effectively in the last few years - caring a lot less about who 'won' and just being interested in 'did we gain?'

Okay, I'll come clean - there's a thing my older brother (Yes, that would be, 'Big Goffy') was involved in and it is freak'n awesome. You'll need your nerd hat on:

https://ai.facebook.com/blog/cicero-ai-negotiates-persuades-and-cooperates-with-people/?utm_source=twitter&utm_medium=organic_social&utm_campaign=cicero&utm_content=video

 

Thanks LG that is very interesting, sometime ago I had training and some experience in negotiations. A tool like this could be very valuable in the future. I sometimes struggled with all the permutations and found the most difficult things to overcome were peoples stubborn and often mistaken belief in their own opinions. We have a lot to thank the Greeks for but the adversarial  method of debate means there is a winner and a loser. Often a better outcome can be achieved through a more cooperative approach.

 

2 hours ago, Demonstone said:

We already have the resources in place, if you count Al Richardson.

Which emoji represents “groan”?

3 hours ago, chook fowler said:

Mark Jackson had that.

I guess it wasn't very sophisticated back then 🙄

3 minutes ago, monoccular said:

I guess it wasn't very sophisticated back then 🙄

Yep, it was steam driven

And here I was thinking it was going to be a thread about our great CHB Anthony Ingerson

Edited January 16, 20233 yr by Demon Disciple

2 hours ago, dworship said:

Thanks LG that is very interesting, sometime ago I had training and some experience in negotiations. A tool like this could be very valuable in the future. I sometimes struggled with all the permutations and found the most difficult things to overcome were peoples stubborn and often mistaken belief in their own opinions. We have a lot to thank the Greeks for but the adversarial  method of debate means there is a winner and a loser. Often a better outcome can be achieved through a more cooperative approach.

 

Having been involved in negotiations for a fair while your premise makes sense.

I used to call it the "grieving period" in the sense that participants needed to replace their anger with a sense of compromise. Unfortunately we are taught to "maintain the rage."

On the other hand compromise leads to a maintenance of the status quo which is perhaps more harmful in many instances.

Now back to the subject... Palace Dees tell us more about the backgammon app you beat. I would have thought it's a mathematical equation like checkers

34 minutes ago, Diamond_Jim said:

Having been involved in negotiations for a fair while your premise makes sense.

I used to call it the "grieving period" in the sense that participants needed to replace their anger with a sense of compromise. Unfortunately we are taught to "maintain the rage."

On the other hand compromise leads to a maintenance of the status quo which is perhaps more harmful in many instances.

Now back to the subject... Palace Dees tell us more about the backgammon app you beat. I would have thought it's a mathematical equation like checkers

Was taught to stay away from "compromise", always a better negotiated outcome where all the parties felt they got something. The classic trying for a "win win". Part of the art is finding what's hidden and what's really important. One of the keys is discovering what are the needs of the other side and acknowledging them. This is where some AI help would be really beneficial a bit like picking up a "tell" from a poker player.

5 hours ago, Little Goffy said:

AI will really kick in as a genuinely useful thing when it begins to be used to support 'mutual best outcome' estimation in things like trade week.

Not so much that it is cleverer or philosophically deep, but an AI would have the capacity to look at all clubs objectives simultaneously without getting a ripping headache, and without personally caring which club looked like they were 'doing best'.

Actually, that second part is an aspect of why Melbourne have pick-traded so effectively in the last few years - caring a lot less about who 'won' and just being interested in 'did we gain?'

Okay, I'll come clean - there's a thing my older brother (Yes, that would be, 'Big Goffy') was involved in and it is freak'n awesome. You'll need your nerd hat on:

https://ai.facebook.com/blog/cicero-ai-negotiates-persuades-and-cooperates-with-people/?utm_source=twitter&utm_medium=organic_social&utm_campaign=cicero&utm_content=video

 

fascinating

i've been trying to get ai platform execution happening for the last seven years in my various roles and it's FINALLY happening in 2023

 

Edited January 16, 20233 yr by whatwhat say what

16 hours ago, DeeSpencer said:

https://www.the42.ie/western-bulldogs-machine-learning-3213937-Feb2017/
 

About 6 years ago 

ML and AI are two different things 

I wonder whether AI would help with score reviews?

(This is not a joke. I suspect AI can identify trajectory changes in ball movement better than the human eye.)

32 minutes ago, La Dee-vina Comedia said:

I wonder whether AI would help with score reviews?

(This is not a joke. I suspect AI can identify trajectory changes in ball movement better than the human eye.)

i was wondering if ai couldn't help the afl make their rules more understandable and easier to adjudicate

but then i woke up

Can't see AI ever being of use on-field in match conditions.

Imagine having to make a decision that took into account the full range of possible movements and reactions of 36 disparate players without 360 degree vision and mind reading technology. Will never be able to match humans.

 

 

2 hours ago, daisycutter said:

i was wondering if ai couldn't help the afl make their rules more understandable and easier to adjudicate

but then i woke up

Many of the recent rule changes seem to have been made by people with no intelligence, natural or artificial. 

23 hours ago, Diamond_Jim said:

With ChatGPT (the latest open source AI educational app) causing rumblings in educational circles you have to wonder how long until AI enters sport.

My guess is that it is probably already used to some degree in NFL to predict the success probability of particular plays.

Could it have application in Australian rules to predict trends in say centre bounces. I don't know if AI can interpret video yet but imagine if it could.

For those who follow these things, humans lost the ability to beat computers at chess in around 2006 and even now you can download programs to your phone that will beat any human.

As a teacher I'm pushing staff to adopt the capability of ChatGPT as a tool.  it's brilliant even with its current shortcoming's in terms of knowledge and ability to process. I can produce new resources and check existing one's very quickly.  

The area of AI it is still considered that humans have to final say in what is produce or acted upon, as systems even AI at this stage only produce what we tell it to produce.  They are taught to look for patterns and act upon them.  This is evident in ChatGPT responses.   

So I would say existing data science techniques with staff over sight will be the norm for a while longer as the datasets.  With experience and knowledge we can see patterns without within the AFL context without needing to rely on AI systems. 

6 hours ago, tiers said:

Can't see AI ever being of use on-field in match conditions.

Imagine having to make a decision that took into account the full range of possible movements and reactions of 36 disparate players without 360 degree vision and mind reading technology. Will never be able to match humans.

 

 

Suggest reading some of the links above 

or Alpha Go