The new wave will be a movement
This new movement will change everything!
Lately, it seems the whole world is speaking about AGENTIC systems or AI Agents or even Autonomous Agents.
It has become clear that AI is not just a a tool we use today to embrace tomorrow, it is a whole movement that will redefine everything that has come before.
It is going to change the way we work, the way we think, the way we structure our future companies, and the way we innovate.
The new challenger
So what are these new AI Agents and why will it change everything?
AI Agents are AI systems that do not wait for instructions or input and can operate autonomously, making decisions based on tools they are given and data they can access, as well as having access to other agents they can interact with.
Traditionally we had two kinds of Generative AI systems:
1. Data retrieval AI, capable of analysing data and providing insights.
2. AI Assistants, capable of handling multi step research and decision making when asked or prompted.
Fast forward to today and we have a third option called AI Agents. They are Autonomous AI systems, capable of acting and interacting on their own, requiring no real oversight and has advanced decision making.
What about communication and interaction?
This new frontier of AI Agents poses an interesting question: How will they communicate with other Agents, and what will this interaction look like?
For a truly seamless experience, AI agents need their own digital authentication and authorization methods – secure, verifiable identities that build trust and streamline online interactions.
Gartner predicts that by 2028, 15% of business decisions will be made autonomously by AI Agents. I think it might be sooner.
But at the moment our digital identity system isn't ready for AI agents.
Luckily There are companies already solving these kinds of questions. They are building and providing us the capability of Authentication and Authorization protocols of Agent to Agent communications in a secure and reliable way.
How do we stay in control?
Another question and a main concern for many people is: how do we stay in control?
Generative AI has the feature of hallucinations, and even though it may be easy to think it has human like behavior, it is and will always be only a predictive algorithm.
To solve this unwanted hallucination feature, the answer today, is to keep humans in the loop, but with the exponential use of this new autonomous AI Agents that will be impossible to keep up.
We will not have the manpower or ability to handle the volume of decisions that would require human intervention.
It will also take away all the benefits of using AI Agents in the first place.
It will severely limit the speed and power of what these Agents can achieve.
Guardrails
The alternative option to human in the loop immediately springs to mind, that of the introduction of guardrails to stay in control.
Guardrails however will diminish the power and potential of the AI Agents.
Guardrails are also static and would be up to each organisation to implement and enforce. This in itself poses a real dilemma.
So how do we solve this dilemma of keeping control and enforcing guardrails?
Agents as a Service (AaaS)
To get to a solution we must think about using Agents as a Service (AaaS) or AI as a Service (AIaaS).
We already have Software as a Service (SaaS) proliferating the business world today.
So, it is no real leap to imagine that with Agent to Agent communication and Auth security sorted out, we can have AaaS where you can rent out an agent that performs a certain service on a per usage basis.
How do we control these AaaS's
Would it be possible to contain or police these Agents as a Service with specially made Guardrail Police Agents as a service?
This is an interesting take on AaaS by having specifically made/trained Policing Agents to impose certain constraints or processes on other AI Agents.
Think about it this way, you could be hiring Policing Agents as a Service, specifically created to police certain pre set guardrails.
This is not difficult to imagine, after all, they are just Agents as a Service (AaaS) with a special purpose.
What will the future hold?
Looking into the very near future, living and dealing with these new technologies are really exciting and a little scary, but what is certain, is that this new AI movement will change everything.
This movement will fundamentally change everything we know about businesses today, about communication, interaction, social and search behaviours and the structure of most things we have become so accustomed to in the digital world.
It will also change the physical world by changing the organisational structure of companies as we know it.
The new org structure will be leaner, with a flatterr structure, while having less employees, with each critical role having a swarm of AI agents working with and for them, making for a more effective, cost efficient business with higher profits.
The only certainty is change, and it is happening so fast, it is difficult to keep up.
Marius Visser
