Alter Ego
As Is
Most (or all) personal assistant software I’ve come across have been attempts to provide ‘assistance’ to you in a ‘personal’ way (what a finding, who would’ve thought of a more meh start to this post). They have all been attempts to replace a human, who would have provided that assistance to you. They are more efficient (at certain tasks) and definitely cheaper than the human alternative.
There are two kinds of use-cases for these:
(a) Work-ish, where the typical use-cases are to provide you information, manage your schedule, answer people on your behalf, etc.
(b) Life-ish, where the typical use-cases are: to provide you recommendations to satiate your hedonistic desires (hey, like a restaurant, music, movies etc.), give exercise/yoga instructions for you, remind you to do things, provide counselling like a therapist, provide you company like a friend, etc.
You already know many work-ish examples, so let’s skip those; you also know a lot of reminder-like use-cases, so let’s skip those too. Here was a good attempt for the ‘friend’ use-case:
Though they called it ‘Replika’, it was still a ‘friend’ who was going to be similar to you; not ‘another version of you’.
It takes the second meaning of alter-ego — the intimate and trusted friend part:
To Be
How about the first meaning? — The secondary or alternative personality?
I wish to have a digital alter-ego who is actually a member of the Internal Family System (refer) within me. Think of it (him) as a facilitator in those board meetings between the various members of the IFS. He would listen to all viewpoints (from the self which wants immediate gratification, to the self that wants me to have a peaceful / bountiful retired life) and pick the course of action for me. He would know me well-enough to nudge me, cajole me, admonish me, threaten me, scare me, etc. at the right moments. If this were to be done by a different human being (spouse, parent, child, friend, boss, colleague, therapist, stranger etc.), the organism in me would perceive to be a case of nagging, annoying, insulting, condescending, and so on. Instead, this facilitator is just me doing it to myself. (Or in His words, me talking to me, haha. Hey, why am I doing this!)
So, a few use-cases of what this gentleman could do:
(a) When I wake up, he will (in the manifestation of residing on my phone), show me things that would make my day better — which itself is subjective /personal, and cannot be generalized into rules or heuristics for everyone. This would involve showing sports results from the previous night only when the team I support won; showing performance of my overseas stocks only when they are green, etc. Now, if something is on fire and I need to address it before I finish my Wordle and the cup of tea, sure, show that, but not otherwise. There is time to do that. If Techmeme has many stories around the same topic (like an Apple event, several earnings results, funding news etc.) he would merge them into a concise summary for me. This would also consider what time I slept, how well I slept, what I ate the previous day, my likely bowel conditions, and the like.
(b) When I do my weekly review/planning, he will (in the manifestation of residing on TickTick app), be the skeptic when I am too optimistic; and be the Pollyanna when I am too glum. For example, when I make unrealistic plans of completing 10 different docs in a day, he would gently (TK: definition of gentle) nudge me with info on how my energy depletes after completing just one doc, and how many meetings I have on the day. He would show me my calendar, and suggest moving meetings that require higher mental bandwidth to times when I am more alert, and so on. If I am overwhelmed and feel glum, hide things that will add to it, (maybe show me a quick cat video), and then move me to tasks that will slowly get me out of the Ugh field.
(c) When I type c within parentheses, and the text editor converts it into a copyright symbol, he will (in the manifestation within my computer) notice that I am getting annoyed, and propose that I don’t use this editor anymore. Just kidding.
(d) Depending on how much of mucking around (aka BXA) I am doing through the day, he will (in the manifestation within my computer) auto-block websites, and so on.
(e) He will recommend appropriate books/apps etc. for my mental health needs at any point in time. And remind me of the learnings from such books, again at the right times.
Do you get the drift? The whole point is — He should make me feel useful, healthy, and so on, by being my better-self and guide me to reach there. Not as another agent outside me. It is entirely possible that I will start hating myself after this, but that I guess would just be advancing that moment in life. Ha.
How will he learn all of these? And how would the implementor of this solution ensure nothing goes wrong? That is the homework here for the ML experts. ;-)
