How will the concept of the Grand Action Model affect the relationship between companies and their consumers?
Some days ago, I told you about the LAM, for its acronym in English Large Action Model, and I explained it to you through the author of the Rabbit R1 himself. its operation and the way to use it through the gadget.
In the last few days, and seeing that we still do not have all the information we would like - although shipments begin in less than two weeks - We are still trying to decipher how the LAM of the Rabbit R1 works through comparisons or analogies.
What is LAM explained in an easy way?
The simplest way I can explain the applicable usage of R1 is this:
Imagine you hand over your device — phone/tablet/account — to someone you know, and What you are asking him is to do a task for you, but let him do it on your behalf using the device you gave him previously.
This person will execute the action that you have sent them with your device. He returns it to you and then presents in detail what he has done., how you thought about the steps in the solution and what it is going to achieve.
After this, you analyze those steps that it has just shown you and your job after the analysis is to see that it is executed correctly and confirm —here the joke of «confirm, confirm, confirm«— to approve the action if you are satisfied.
The Rabbit R1 is that «someone» to whom you gave those capabilities to do the action for you. To whom you have given permission to execute those acts on your behalf. The Great Model of Action is how that someone thinks.
Given the there is no real precedent for something like a LAM action model It will be interesting to see how consumers, businesses, the market and even regulators respond to that change with new layers of execution.
Let us take into account that the Google Assistant and Siri do not count because they are ecosystemic. R1 is being executed outside the ecosystem of the actions it performs.
Companies and the new approach that the Rabbit R1 presents to them
Well, taking into account that definition of LAM and how someone acts on our behalf, it seems that an interesting melon is going to open when it comes to dealing with different applications or web pages that are outside the Rabbit R1 environment.
When we install an app or even register on some websites —such as social networks—We accept the so-called Term of Service, which are ultimately the conditions that that application sets for its use.
The fact is that when you accept them, what you do is sign an agreement by which the use you make of it is reflected and done by you, as an individual user of that application. What happens if the executor is no longer you but a LAM?
Being a technology new we enter into a new need to correct and change the way those Terms of Service worked, since they are not compatible with what the Rabbit R1 promises a priori.
Rabbit R1 will work in such a way that all requests will go through a "virtual machine" with a series of specific IPs for all the owners of the gadget, which could lead to a couple of possible options:
- On the one hand, companies will want you to spend time on their App or website, so executing that action without the involvement of the end consumer is not liked by the vast majority of companies that want their business to work.
- On the other hand, there are tools that detect suspicious connections per user based on session activity associated with the approximate location of the IP. And they immediately automatically block a user's account based on this type of behavior, until its authenticity is proven.
All this implies that we have to look for new alternatives when this technology is implemented, since at the moment it is unexplored terrain on which we will have to learn from scratch.
In principle, the action process would be as follows:
OFF DEVICE (R1 and user)
- The user asks R1 for an action.
- R1 knows the interface, events and how to "act" on each of them.
- A decision is made about what and how a solution to a problem should be executed.
- The user authorizes R1 for the action.
ON THE DEVICE (R1 and user):
- Those instructions/events are executed step by step.
- The problem is resolved and the process ends.
As you can see, it is a difficult issue and we are really going to have to face quite a few problems when adopting this technology, especially on the part of the websites or applications for which we want to educate our R1.
