Ultimate Personal Assistant

Projects Suggest a Project
1

This is not a totally new project idea, but I hope that it might help enrich some others that are out there.

Context: technology that is “pushed” needs marketing effort to convince the target User of the values. it is rarely holistic (examples abound that damage the environment; require admin, support and hassle to make them work; exposes the User to fraud, unfair pricing etc. etc.).

My desirable User Experience may be as follows:

  • give me a PA that costs nothing at first, does nothing and does not put me at any form of risk or disadvantage (download from AI marketplace)
  • a process develops whereby the PA gets to know me - in confidence, what I like, dislike, hold dear, do on a regular or irregular basis, and - most importantly - any critical characteristics of my existence (e.g. severe peanut allergy). All this data is secure and trusted
  • at the same time, I get to know the PA - what they can/not do, and how well. I need to understand the User Experience: if I subscribe to a particular service, how will it work for me, what are the benefits and opportunities? (This could be simulated for me to start with)
  • There will be a hierarchy of services… imagine a PA that grabs discounts for particular products and services that I like, recommends new or better alternatives, and over time automatically harvests those discounts and gets me the “best deal in any context”
  • I can select-subscribe-deselect from the PA services at will. I might activate the PA “finder” function, and ask them where my keys are, or where the nearest open pizza restaurant is (and then my PA has a table ready and waiting)
  • the services can be as granular as required, and follow a trusted update cycle as the capability develops. My PA would (of course) integrate with AI home assistants, IoT systems etc., and the levels of action are guaranteed to be in line with the level of trust I have in the PA. Maybe it starts with shopping, then security, then on to medical care and life planning…
1 Like
2

Thank you @PhilW, I think that is what we are doing:

Sounds quite supportive :slight_smile:
More on this:
https://aigents.medium.com/

3

Thank you for sharing Anton, there is certainly a high degree of overlap!
I particularly like your description of How it Works (s.8) and How to get started (s.15). The latter implies that the Agent can be infinitely tailored to the needs of the specific User, and the content could be updated in real time as new functionality is developed.
Very best of luck with your commercialization, Regards PhilW

1 Like
4

It may not be much longer before this project suggestion becomes a reality:

However, I read phrases like "AI…with really good long-term memory”, “(that) will encourage users to invest more time and share more of their personal history with them” and with predictions about security, privacy, bad advice, wrong information, inbuilt bias etc.
One defence is that “many people’s privacy expectations have changed”. My expectations have not changed, maybe there is an external force trying to change them?
With billions of dollars flowing in, and with proposals for “opt-in” web page observation etc., I feel that we may be losing the opportunity to make the User Experience I imagined above truly decentralized and beneficial to all.