Personalization has been a human endeavor for so long. The web evolved from the same webpage for everyone to each video/post recommended based on your activity. AI is also going through the same phase; from the similar response for everyone (2022-2023) → memory/RAG (2024-now) → deep personal understanding/proactivity.
This dream of personalization has been chased forever. It started with newspapers, when we began showing news related to region with different publications for each area. With radio and TV, this became more accessible, as we could customize to even smaller regions. With the internet, we followed the same pattern; starting as newspapers, the same webpage read by everyone, then region-specific websites, then translating based on user location.
With mobile and advancement in data and user understanding, it became more personal. Now we have a device which is always with us. This led to Uber, Instagram—more personalized and more aware.
With AI, we are redoing this whole thing again (as we did with web). AI started as a chatbot with the same answers for everyone irrespective of the person (asking the query); think early 1994 web. Then in 2 years came memory and RAG, now AI has a prompt injection of what's relevant to user and what it has done before; think pre-mobile web with user logging and recommendation, like Facebook and YouTube.
The next change will be the mobile phase of AI. It will be more personal and tuned to the user always together. AI will have the benefit of deep computation and reasoning on top of that, leading to hyper-personalization of user/system interaction.
We at Intelligence Company are a applied research lab focus on understanding "us" and ready with whats next always!
Vikash & Harsh