▲ ▼ Make decisions for me
Every single day we make several decisions with varying level of consequences. I would like help to make faster, thorough decisions which has better probability towards favorable outcomes.
Although technology influences several of our decisions, I wonder whether technology is at a stage where I can ask a digital assistant to make a decision for 'What dress should I where today' to 'Which insurance policy should I buy' considering all requisite personal data to make a decision regarding these i.e. Calendar, Health, Finance etc. are already with the Big Data companies; Only difference being the decision should be in my favor instead of an advertiser.
I would divide all decisions into basic everyday decisions and important decisions that will impact the rest of your life.
For the simple decisions, like which dress you should wear today, trust your intuition, common sense, randomness, or spontaneity.
But for the long-term decisions, you need to evaluate how they will affect your life from different perspectives. For that reason, I would suggest using the online tool I built, https://www.1st-things-1st.com
For example, if you want to choose an insurance policy, you would have to do a little research and use these criteria to think through:
- How much money would you be willing to leave to your loved ones if you left this world suddenly because of an accident?
- How affordable is it?
- How flexible is it?
- How well-known is the company?
- How much do you trust the company?
You would list out the insurance policies you are considering and would evaluate each policy by each criterion in percentage. "1st things 1st" would calculate which of the insurance policies fit best to you.
There are also some prioritization templates to start with, like choosing a home, university, car, or immaterial things like lifetime goals, daily activities, ideas, etc.
Are you sure this is a technology problem? I agree that any technology platform is going to favor whoever is paying for the technology. So I agree that the big data companies aren't your friends in this regard. On the other side, the decision about what color dress to wear today is a very insightful one to consider. Since how we dress affects our self-image and our emotional state, and it depends at least partially on weather, culture, social status, tasks to be done today, expectations of who we meet today, what we want out of those meetings, and many more things, it's a hard problem, probably too hard to technology. Tech can't know that you're feeling a tad self-ware today, and so you need to wear that power-suit that got you compliments that one day last month.
Instead, I think this problem is one of personal cognitive load and attention budgeting. These are nicely addressed by various philosophies and lifestyle choices (think Buddhism, Danish Hygge, Ancient Greek Stoicism, the modern Simplicity Movement as just some examples).
My take on this problem is classify it into a couple of categories: 1. recommendations for things you only really need to do once (buy insurance, choose a college to attend, etc), and 2. things you do lots of times (eat breakfast, do chores, etc)
Things that you only do once might merit more thought per instance, whereas things you do repeatedly might be easier to do more quickly due to having had more experience or more data working with them.
For the first problem of having unrelated decisions to make, I don't have much advice other than that if it's possible to classify things in a way to make them slightly more repeated then that could be helpful. For example, buying insurance could be treated as an instance of buying something, and when buying something it could be interesting to know the value of your time to inform how much time to spend choosing.
(I'm encountering a character limit so I'm splitting my response into a few comments)
Yeah I would like for more suggestions to be more tuned for me and less tuned for advertisers too.
If there were a project that was not monetized but that requested some feedback from you to help make itself better, and whose purpose was to help to quantify and make informed, fulfilling decisions about some things, would that seem worth checking out?
I think anyone claiming to solve this should actually make money, else people won't trust their data with them. But to make money they should solve this first or at least showcase the ability to do so, I wonder how that can be done.
Maybe reputation in ML/DL field can help @Abishek ?
For the problem of quickly making many similar decisions, I agree with Abishek about machine learning: I think it can provide the opportunity to be faster, more rational, and eventually more explainable.
How would you quantify what you're trying to optimize? Do you think you could recognize a satisfactory decision after making it, potentially assigning a happiness number to it, to inform a machine learning model?
>Only difference being the decision should be in my favor instead of an advertiser.
Well articulated. Manipulating user decision in favor of an advertiser is easier because advertisers want just one outcome (conversion) where as individual decisions for a positive outcome might depend upon numerous variables and defining positive outcome itself would be a huge task.
Looking at where Machine Learning/Deep Learning tech is heading to, A digital assistant which can take real-time decisions for us could be inevitable. But those who have large swathes of data on us are usually into serving advertisements and so the problem of data collection to serve our own interests should be addressed first.