News Bob Iger Outlines Disney's Growing AI Strategy

[Basil of Baker Street]

This will end up like analytics in baseball...

 

[HMF]

They're going to pay less people to operate stuff, less artists to make stuff, and yet prices will keep going up while the experience gets worse as a direct result of the aforementioned less people and less artists.

I do think the need for less people to do certain jobs and make things more efficient is a good thing. I do not want it replacing artists and creative people for obvious reasons. Disney isn't really a "Model Employer" anyway.

 

[Tom P.]

AI is so horribly, horribly bad at just finding and reporting basic factual information that I can't believe anyone actually believes it will be as revolutionary as claimed. It often literally just makes crap up and presents it as solid fact. If it cannot handle basic tasks that a search engine could get right, how is it going to revolutionize everything, let alone Disney?

As Kevin Flynn said to the Master Control Program, "How are you going to run the universe if you can't even answer a few unsolvable problems?"

 

[Trauma]

AI is so horribly, horribly bad at just finding and reporting basic factual information that I can't believe anyone actually believes it will be as revolutionary as claimed. It often literally just makes crap up and presents it as solid fact. If it cannot handle basic tasks that a search engine could get right, how is it going to revolutionize everything, let alone Disney?

As Kevin Flynn said to the Master Control Program, "How are you going to run the universe if you can't even answer a few unsolvable problems?"

AI is in its infant stage. It could have value in specific use cases within the Walt Disney Company.

However the way it’s being used now is just a buzzword.

I still think the next 10 years will see some very interesting changes.

Personal robots for use in the home will be coming soon. People will laugh and call them stupid because they will fail at the most basic of tasks.
However I wouldn’t be surprised if 15 years from now all housework will be done by them.

My point is tech continues to advance and just because AI doesn’t seem very useful currently, doesn’t mean it won’t evolve into something that is.

 

[BrianLo]

It’s surprisingly useful in medicine workflow, though arguably fixing a problem medical records created. Though we are not asking for any problem solving or decision making. It is quite capable at listening and reproducing a surprisingly good note.

 

[John park hopper]

Bob's vision of WDW -----virtual reality --there will be no physical parks just rent a pair of these. Look at the $$$$ they will save and make (sarcasm I think)

 

[MR.Dis]

This will backfire, when the new CEO comes in next year (as expected), either if it’s Josh, Dana, or whoever; one of their first moves they do is to kill off any and all AI Generated programs at the company

Dana coming from the studio sector will definitely kill AI to appease the actors/writers of the Hollywood crowd. Not so sure about Josh as he might see how AI could help him save millions in labor costs at the parks. As I have said before, not a fan of either of these two (I preferred Dana until she messed up the handling of the Kimmel suspension). I just do not believe there is this golden person waiting out there who can swoop in and run this very highly diverse corporation with any competence.

 

[Disney Irish]

AI is so horribly, horribly bad at just finding and reporting basic factual information that I can't believe anyone actually believes it will be as revolutionary as claimed. It often literally just makes crap up and presents it as solid fact. If it cannot handle basic tasks that a search engine could get right, how is it going to revolutionize everything, let alone Disney?

As Kevin Flynn said to the Master Control Program, "How are you going to run the universe if you can't even answer a few unsolvable problems?"

AI is in its early days, to use your search engine analogy it is like if you were trying to compare search today to where it begin in the early 90s. In the early 90s it couldn't bring back even the simplest results that we take for granted today. It improved over time and got to where it is today, which by the way is backed by AI for most search engines.

So give it some time, its already improved leaps and bounds over where it was even just a few years ago.

 

[DrStarlander]

Personal robots for use in the home will be coming soon. People will laugh and call them stupid because they will fail at the most basic of tasks.
However I wouldn’t be surprised if 15 years from now all housework will be done by them.

I do wonder though, as humans are prone to anthropomorphizing, that if there is a humanoid-like (physiologically, voice, responsiveness) robot in the house and all it does is chores, if that will feel good to people. I know people have had human "servants" in their homes for millennia but imagining that myself I think I would feel guilty and lazy! And that thing better not be walking around in the dark at night cause that would give me a heart attack.

 

[lazyboy97o]

I do wonder though, as humans are prone to anthropomorphizing, that if there is a humanoid-like (physiologically, voice, responsiveness) robot in the house and all it does is chores, if that will feel good to people. I know people have had human "servants" in their homes for millennia but imagining that myself I think I would feel guilty and lazy! And that thing better not be walking around in the dark at night cause that would give me a heart attack.

People already anthropomorphize and thank the robot vacuum that’s just a giant hockey puck.

 

[HMF]

I do wonder though, as humans are prone to anthropomorphizing, that if there is a humanoid-like (physiologically, voice, responsiveness) robot in the house and all it does is chores, if that will feel good to people. I know people have had human "servants" in their homes for millennia but imagining that myself I think I would feel guilty and lazy! And that thing better not be walking around in the dark at night cause that would give me a heart attack.

What about the people who literally can't do chores or housework because of disabilities. To them AI will be a godsend. I can't drive so I know it will make my life a heck of a lot easier.

 

[HauntedPirate]

I'm sure Bob and Josh's ultimate goal is to create a LLM that tells you how great of a time you had at WDW, complete with computer-generated pictures and videos, all for a $10k credit card charge.

 

[lazyboy97o]

What about the people who literally can't do chores or housework because of disabilities. To them AI will be a godsend. I can't drive so I know it will make my life a heck of a lot easier.

But does that require the sort of generative AI that is currently hot? Waymo is a lot older than Gemini.

 

[HMF]

But does that require the sort of generative AI that is currently hot? Waymo is a lot older than Gemini.

To be honest I am not really tech savvy enough to know what you mean by this.

 

[lazyboy97o]

To be honest I am not really tech savvy enough to know what you mean by this.

Waymo, Google’s self driving car, has been out on roads for years now and works surprising well a good chunk of the time. Gemini, their new AI product is hot garbage that hallucinates all sorts of nonsense. They didn’t need “AI” as is currently being hyped to make self driving cars.

 

[UNCgolf]

A lot of people don't realize that the "AI" being touted right now is not really AI at all, at least in the way people generally think of AI. It's more or less a predictive algorithm that "guesses" (for lack of a better word) the next word in a sentence based on how it's been trained.

It's also arguably getting worse (there were indications that the most recent ChatGPT was less accurate than the prior version), and that might accelerate because more and more people are using it to write things, and then that LLM generated content is being fed back into the LLM to help it generate more content.

There's also the separate issue that most people won't be able to actually afford it when it comes time to pay. It's astronomically expensive to run and the companies can't keep footing the bill indefinitely; last I saw ChatGPT would require something like a $5k yearly subscription fee from each user to even approach profitability.

Then there's video/audio/"art" generative AI, but that's whole other thing with a host of copyright and IP issues.

 

[Disney Irish]

last I saw ChatGPT would require something like a $5k yearly subscription fee from each user to even approach profitability.

OpenAI will make a majority of their profit on ChatGPT from enterprise licensing rather than direct subscription fees from individual users. So that'll actually make any subscription much cheaper.

 

[HMF]

Waymo, Google’s self driving car, has been out on roads for years now and works surprising well a good chunk of the time. Gemini, their new AI product is hot garbage that hallucinates all sorts of nonsense. They didn’t need “AI” as is currently being hyped to make self driving cars.

That is good to hear but there are people who have much more problems than I do that would really benefit from AI. I understand the technology is in its infancy but I can see that it might be useful to many people in the future. I do worry about the negative aspects too especially when it comes to the realms of the arts.

 

[HMF]

There's also the separate issue that most people won't be able to actually afford it when it comes time to pay. It's astronomically expensive to run and the companies can't keep footing the bill indefinitely; last I saw ChatGPT would require something like a $5k yearly subscription fee from each user to even approach profitability.

I do think profitability will not matter so much in the future. We are limited in our understanding because of the world we currently live in but like with Carousel Of Progress, I am sure people who lived in the 1890s could not fathom life in the 1960s. We don't know what massive technological or social changes will happen within the next decade even.

 

[UNCgolf]

OpenAI will make a majority of their profit on ChatGPT from enterprise licensing rather than direct subscription fees from individual users. So that'll actually make any subscription much cheaper.

I'm not sure that's true. Private businesses are generally creating/hosting their own internal versions of AI software, because they don't want their data available to the public. Obviously they need the assistance of a company like OpenAI to get it up and running (although I think most are using Microsoft or Google), but the individual user version of ChatGPT (and competitors) will essentially be a different product that accesses different information. It's not like the individual user version will stop costing astronomical sums to run; I don't see enterprise use generating enough money to cover the cost of millions of people using the public version.