Hi Lisa! I love your videos- I feel like I am learning so much and I really appreciate that you have been trailblazing these changes as they occur in real time and making the knowledge accessible! Would you consider making a video demonstrating a set-up in which multiple custom Copilot Studio agents are connected?
I enjoyed this balanced discussion. Yes, there is a lot of hype out there. And much to come in the future. But there are a good many things we can do today.
Hello Lisa ! Just discovered your TH-cam channel today and this is great ! Thank you, I have a few questions, and sorry if you have already answered it somewhere: - All agents seams to be very linked to the user that created them, can we easly "Transfert" the ownership of an agent from an employee to an other employee ? - Is it possible to use service account to create agent so that it is not linked to somebody ? - We saw that the email that your agent send has been send by taking your Exchange Online account, is it possible to use a service account for that ? So that it will be a "Nameless" company user rather than an existing person - We can imagine that Agent use standard OAuth 2.0 authentication, does this means that I will have to log in every 90 days in my agent to refresh the authentification token ? - (This one is a stupid question) In the trigger option "When an email to" what email address can you put in ? Only yours ? Can you put your CEO email ? (lol), a service account ? a distribution list ?
They are just nothing but Micro Services with RAG with 3rd party API connections . Agents just respond to Natural Language , where as Non Agent applications responds to user requests or API calls
I think he is making stuff up and trying to run interference. He has also said you can’t even find Copilot on the Microsoft website which is simply ridiculous. He claims Microsoft aren’t using it themselves, I work with Microsoft people every day and see them using it.
@@LisaCrosbieThank you Lisa. I think his interviews are good for Microsoft to stay on their toes and also tame down some of the expectations people have from Microsoft.
Agents are not a new concept - but they have matured to a level where they are actually reliable. But in Microsoft Co-pilot not so much. Nice framework MS has built but probably the worst place to build it.
So life is going to become increasingly impersonal. A world run by bot agents. No human interaction. Just spock like logic. A cold lifeless world where we all die of loneliness.
Most information workers spend a fraction of their time connecting with people and a large portion ‘navigating and managing processes’. Agents serve to rebalance that equation, allowing people to focus on high-value and personable activities. For self-serve customer interactions, it means not waiting in a queue filled by trying to solve simple queries, and being escalated to a human agent when required. It gives capacity to organisations trying to scale.
Hype, overhype probably, like most of AI tools. Companies will rush and inflate and burn their budgets just to realize that it was all waste of time and effort, and that they're left out with huge tech dept to be paid. AI is a bubble, tech companies are constantly advertising it as revolution bullet train that won't stop and you either join or die. that in of it self should be a red flag to any company. if it is advertised like that, everywhere even in your damn fridge, then someone is trying to milk you. there is hardly ever a use case for it, or at leas a real use case that can have ROI. I did a couple of AI use case Analize for multiple companies, and they age eager to burn a stock of cash, but to find really good and useful case for it, that would also meet all security measures and will payout is like finding unicorn. To much hype, to many startups offering solutions for all your problems. I know that high managers love to buy this stories but there is simply no demand for it from workers it is forced by managers who ignore all the red flags around them. Believe me when I say it will burst, and burst it must and so man will freakout and loose their jobs.
Hi Lisa! I love your videos- I feel like I am learning so much and I really appreciate that you have been trailblazing these changes as they occur in real time and making the knowledge accessible!
Would you consider making a video demonstrating a set-up in which multiple custom Copilot Studio agents are connected?
I enjoyed this balanced discussion. Yes, there is a lot of hype out there. And much to come in the future. But there are a good many things we can do today.
Thank you, really value and appreciate that from you.
Hello Lisa !
Just discovered your TH-cam channel today and this is great ! Thank you,
I have a few questions, and sorry if you have already answered it somewhere:
- All agents seams to be very linked to the user that created them, can we easly "Transfert" the ownership of an agent from an employee to an other employee ?
- Is it possible to use service account to create agent so that it is not linked to somebody ?
- We saw that the email that your agent send has been send by taking your Exchange Online account, is it possible to use a service account for that ? So that it will be a "Nameless" company user rather than an existing person
- We can imagine that Agent use standard OAuth 2.0 authentication, does this means that I will have to log in every 90 days in my agent to refresh the authentification token ?
- (This one is a stupid question) In the trigger option "When an email to" what email address can you put in ? Only yours ? Can you put your CEO email ? (lol), a service account ? a distribution list ?
They are just nothing but Micro Services with RAG with 3rd party API connections .
Agents just respond to Natural Language , where as Non Agent applications responds to user requests or API calls
Would love to know your thoughts on Salesforce CEOs opinion that copilot agents are a failure and that they are not being used in real world.
I think he is making stuff up and trying to run interference. He has also said you can’t even find Copilot on the Microsoft website which is simply ridiculous. He claims Microsoft aren’t using it themselves, I work with Microsoft people every day and see them using it.
@@LisaCrosbieThank you Lisa. I think his interviews are good for Microsoft to stay on their toes and also tame down some of the expectations people have from Microsoft.
you'd probably be amazed by what o3-Mini can do then (when o3 Model gets released).
Thanks for the tip, I’ll keep an eye out.
So what percentage can you estimate will Agents replace people.
I think that will vary a lot depending on the types of roles, and while we will see some jobs replaced there will also be new jobs created.
Its basically office job killers. Most of us will be working manual labor again or on govt assistance soon because there will be no basic jobs left.
Are AI agents the new RPA?
Basically the same vendors rebranding for the hype - AGI is the main goal
😮
Agents are not a new concept - but they have matured to a level where they are actually reliable. But in Microsoft Co-pilot not so much. Nice framework MS has built but probably the worst place to build it.
So life is going to become increasingly impersonal. A world run by bot agents. No human interaction. Just spock like logic. A cold lifeless world where we all die of loneliness.
Most information workers spend a fraction of their time connecting with people and a large portion ‘navigating and managing processes’. Agents serve to rebalance that equation, allowing people to focus on high-value and personable activities.
For self-serve customer interactions, it means not waiting in a queue filled by trying to solve simple queries, and being escalated to a human agent when required. It gives capacity to organisations trying to scale.
Well, may be for you
Hype, overhype probably, like most of AI tools. Companies will rush and inflate and burn their budgets just to realize that it was all waste of time and effort, and that they're left out with huge tech dept to be paid. AI is a bubble, tech companies are constantly advertising it as revolution bullet train that won't stop and you either join or die. that in of it self should be a red flag to any company. if it is advertised like that, everywhere even in your damn fridge, then someone is trying to milk you. there is hardly ever a use case for it, or at leas a real use case that can have ROI. I did a couple of AI use case Analize for multiple companies, and they age eager to burn a stock of cash, but to find really good and useful case for it, that would also meet all security measures and will payout is like finding unicorn. To much hype, to many startups offering solutions for all your problems. I know that high managers love to buy this stories but there is simply no demand for it from workers it is forced by managers who ignore all the red flags around them. Believe me when I say it will burst, and burst it must and so man will freakout and loose their jobs.
"Internet is just a passing fad" (Bill Gates)
@@VurtAddicted No, AI hype is more like DotCom bubble around the year 2000. Nobody denies AI is useful and will stay. But is is overhyped now.