Agenda is nothing but automation 20 with memory and side effects. For example, the old school automation will go and book the tickets as per the availability of the buses. But the age and decay can look at your calendar, can look at your past bookings and then look at the weather pattern and decide it is based on your cost or time and then book the tickets. So right now platforms and frameworks are there to build the agent. OK, but it is still in the developer level. First 12 to 18 months, it will be more of an experimental thing. After 12 months, this would become a common terminology and a lot of people. Will start selling the agents so that the consumers can buy agent. For example, I can buy a agent for posting on social media. I can buy a agent for replying to e-mail. Upper level CXO people used to have personal assistance. Now a common man or an individual seller can also have the personal assistant with the agent. One agent can look at the data which is there and tell what is the demand. Another agent can go and reply to the customer e-mail. This would be very useful for the SMBs and budding entrepreneurs. The important factor that will drive the adoption of agent picks two things. One is the cost, second one is the accuracy. So one of the key things that needs to be done is can you connect a multi modal stuff? For example, right now we have a model for language, we have a model for vision, we have a model for audio. If we can connect all the different models in a seamless way and this would be the dominant feature. It could be more than the humans which is out there in the problem solving domain. So think like your old school automation or the current RPA is a train which only goes in a particular track whereas the agent K would be a self driving car, the same automation. But it can make intelligent decision and it can drive accordingly.
