Topic: [POLL] Using AI to predict outcomes - page 7. (Read 10042 times)

Asking for a full prediction would be industrial usage and they probably dont want people to do that even if it could be done with zero further work required.  I bet its possible but we'd have to guess on the accuracy of it all, perhaps that does not even matter so long as the machine can self examine the results and reasoning to then improve future bets.    I dont reckon AI is quite that good in terms of self improvement.
  Easiest way to get an answer from AI is ask for a story but then it might just say anything.

If you run enough queries you'll probably have lots of different results
those who think AI can predict things like that probably aren't familiar on how LLMs work, at least at this points an LLM is a bit more like a good model to select words according to the best mathematical/statistical response

At the end of the day, any casino will only offer bets that include some benefit to them, as in, the odds are against the gambler, and in favor of the casino.

It doesn't really matter how good the AI can predict the outcome, the odds are already calculated so that the casino will end up winning more than the gambler.

And the thing is that casinos are probably already using the most advanced AI models to generate their odds.

I guess it will never happen. The word "sport betting" refers to a game or gambling wherein real people are involved; they play sports like basketball and such. How do you think AI will predict who will be the winner? AI doesn't have the ability to predict the course of the game as it involves humans, which AI can't control, so it's nonsense that some will rely on it. Maybe AI could predict the winner. That's because there are only two winners in two teams; it is 50-50! I don't understand why some rely on AI; they don't think what the only capabilities of an AI are. Unless AI can control or manipulate the course of the game, that's when AI would be realiable in terms of sports betting. I'm not against AI; don't get me wrong, but I don't think AI has the capability to outsmart or predict anything that humans do.

Interesting. I have no idea about that too. Maybe someone here can explain it to us? Or maybe run some experiments and show us the results. I personally think that it's not about AI's inability to pick the right data. It's about AI's refusal to do so, for some reason. But maybe I'm wrong here. Maybe it's possible create a working model. Interesting.

Well, unfortunately I haven’t been able to get AI to flat out predict the winner of sporting events yet. Sure, there’s ways around it with nonsense, but I’d really like to get the AI’s genuine opinion of who the winner of a sporting event will be. I guess this is still a no-no in the world of AI, but I’m certain it’s only a matter of time.

nice betwrong
I thought so

you can probably still use LLMs/AIs powered with something like Wolfram and data models to check past statistics and create a working model that is profitable over the long run
I have no idea on how to do it but I could see this happening.

Good question. Now, after more tries, I'm inclined to believe it was just luck. My subsequent experiments with Bard AI have shown that it's not much of a use for predicting the outcomes of the sports events. For example, Bard refuses to take into account which players are injured, although I said where to look for the info, https://www.premierinjuries.com/injury-table.php. And it predicts 2-1 in favour of the strongest team almost always, and that is not good enough, obviously.

what makes you think this wasn't just luck but something taht is replicable and you can implement in a profitable way over time?

congratulations on the win and I'd be really happy to lose 8 out of 10 if my 2 wins got me a huge multiplier like that, but remember this could have been just luck, like a complex looking RNG asnwer created by an advanced LLM, but still no different than throwing dice.

At the moment there is no reliable way to determine if a text was written by AI or a human.

In the same way, there will probably be no way of determining a bet was the output of an AI or a human intuition.

Casinos will probably will have to keep an eye on someone winning too much money, but they won't be able to know why, and all their checks would come back reporting that the gambler has done nothing wrong.

I have never thought of this but it may as well be very much possible after some time and of course no human being can check that many resources in such a short amount of time,it can help make better decisions when placing live bets and that could be a good start for AI helping humans winning a bit in sport betting.Of course that AI would be used by the casinos too and they would put mechanisms in place to protect themselves by changing the odds faster and thus leaving many of us without the possibility of betting.

I think that only when AI will be developed to check a lot of resources before a game starts and make a decision that statistically will be the best,that would be a good start.

That's what I thought. It's all about a perfectly written prompt, right? My failed experiments with AI predicting outcomes don't prove anything. This was only a first try and I didn't think of a good prompt that much, I just wrote the first thing that came to mind. Yes, AI was wrong 8 times out of 10, but the same AI helped me to win 22x of my bet



I think I will be experimenting more with it. Maybe AI plus with some human corrections is the answer.

I have always thought about something, when a technology is being born it is harmless, it cannot do much, it has the acceptance of everyone, even children themselves, they can manipulate it, but we cannot allow ourselves to be amazed by what man can achieve, in In the middle of everything, we ourselves can be creating the aram that will be our executioner, and then we cannot ask for mercy, because things are like that, every time AI advances, there will be many threats, not for now, there are many that are in the stage Beta, they have errors, their information is extracted from the internet directly, things are not as they should be, it is something common, personally, right now I would not worry, but later I will, because there are more trainings of the robot, there is more time, the algorithms are more perfect, it could be that the robber has much more logic, and that is where the bad thing begins, by having more logic he thinks better than Natural Intelligence, and that is something that does not We must allow, but control, an AI must be controlled, because at the least expected moment it will control us, and that is not the idea, AI has advanced, yes but to a certain point, and those advances that for me are nothing, many people They marvel, many want to be in the spotlight and be participants and protagonists.


Enthusiasm is not bad, it's good, the bad thing is as time goes by, things with AI will develop more, as I said, there may already be robots that calculate at least the end of the number pi, something that the being human under his own criteria and tools has not been able to give, however this for an advanced AI it will do it in 3 seconds, so this is the only thing we must take into consideration, it will beat us in speed, and that is not good, it will come in that an AI will handle many jobs, it will take possession of many high-care objects, because since it is unbiased, they have no emotions or walk, then man will believe that he can trust.

I think that at the end of the day AI should be able to "watch" and process those practice videos and form an opinion just like how a human would do.

AI can also integrate information from many other sources and watch multiple video streams faster than the time of the videos, which would be impossible for a human.

I think AI has the potential to be better than humans in pretty much every area, eventually

I'd say it depends, I saw this guy who did a model that removes all bias from decision, only uses data
and the model is overall profitable, I think he's up more than 70% in some months or something like that

so, there are ways to do it without emotion/human decisions.

Its interesting that a lot of people believe sports betting is just about data, but there's more to it than that, isnt it? The human element of the game is being lost if we only use AI to make predictions. Lets say a star player is injured and the replacement is a beast in training but has little to no internet statistics. That would be completely lost on AI, wouldn't it?

You make a valid point: AI might not be able to discern subtleties that humans do, such as the atmosphere created by instructional videos. Bookies will surely realize the significance of these "hidden gems" if we humans do. They may be ahead of the game, adjusting the odds in light of these revelations. Thus, perhaps, just perhaps, in the realm of sports betting, AI and human intuition go hand in hand.

AI's future development will definitely be much better so that AI can be truly useful for humanity and can help human activities. That is the goal to be achieved so that humans can benefit from the presence of AI. But for now, we have to wait for the developments carried out by each developer, especially for AI which will later be used to guess the results of matches. Maybe AI can calculate a team's win rate percentage based on previous matches. But now, the results may not be as accurate as we want so we still have to look for additional information.

Yes, it is almost like a second, or even third wave of amazing technology coming to people and many are left behind.

It is like they say, future is already here, it's just that it is not evenly distributed.

In some countries or areas you feel like living in the future while in others you still feel like stuck in the past.

Techonoligical advancement is something that inevitable but we shouldn't really be that hope for those things to happen soon yet there  would really be tons or industries that would really be mainly be affected once quantum computing would really be already that available or has been created. We know that the current ChatGPT did really make that entrance on which it did really hype up the entire world about its usage and relevance but of course there would really be those assumptions that had been made on.
It would becoming more better and better as the years passing by on which advancement would really be  there and its something inevitable.
We dont know on how far iwould really be able to reach out but we know that as long that it would really be continually to be developed then expect
that it could really be able to be more useful to humanity but the fact that there would really be some cons on some corners or things but doesnt matter
which as long it would really be outweighing those risks by its utility then this what matter the most.

we don't need quantum computers or going so far to find good prediction models that can have an edge against the house in sports betting
there's people already creating models like that using data science and programming

there's lots of data available for sports and people can use this data sets to study profitable ways to bet

I saw some people doing so in instagram this days like leandrowar