Hacker News new | past | comments | ask | show | jobs | submit login
Show HN: Condorsay, a decision-making assistant with help from GPT-3 (condorsay.com)
6 points by dfrankow 2 days ago | hide | past | favorite | 7 comments

First it makes a chat thing pop up uninvited, taking up 1/4 of my mobile screen and covering up my input and the suggestion drop down.

Then I get a list of things I don't want to filter by.

Then I get this:

> We're glad you're here! In order to continue making better decisions, please login with a Google account by clicking the button below. We'll take you right back to your decision once you do!

Nope nope nope.

Thanks for your feedback.

We'll look into making the chat less obtrusive. I also don't love aggressive chat boxes, but didn't realize Intercom did that on mobile.

it is quite laborious. Asking me question after question and there is more than 70 questions. I gave up after clicking 10 of them

Hi HN.

I'm Dan, CTO at Condorsay. We've built a tool that helps you make decisions, alone or with others.

* Summary

The short version of our tool:

- pick a goal (something to decide)

- pick factors important to the decision (helped by GPT-3)

- pick options (helped by GPT-3)

- use pairwise ranking to learn what you or a group think of the options

- see the results, including text notes (if you made the decision with others) and dissenters (people who disagree with the group)

NOTE: it does require a Google login to get past the "factors" screen. However, you get 5 decisions for free, so you can do everything for free. I read login makes HN cranky, but that is how our tool works. (A decision has to be owned by a user in the DB. Also, we are protecting against someone burning tons of money on GPT-3 calls.) Please don't be cranky.

* Motivation

We believe that decision-making could be improved by a tool that puts structure around it.

Some benefits:

- clarity: structure and record your decisions, alone or in a group

- focus: force hard choices with pairwise comparison

- revealed preferences: you actually don't always know what you think, but you learn through the simplest possible gut-level calls (pairwise choices). By the end, the results make sense, even if they aren't what you thought at first.

Further benefits if you’re making a decision in a group:

- alignment: get a group of people on the same page, by having the most important discussions quickly (i.e., where people disagree)

- independence: express your preferences before you see anyone else’s, to avoid information cascade

- asynchronicity: coordinate people in our increasingly remote-work world, mobile-friendly

Our scoring is straight-up Analytic Hierarchy Process (https://en.wikipedia.org/wiki/Analytic_hierarchy_process). For AHP in Python, we use the lovely https://github.com/PhilipGriffith/AHPy.

Happy to talk about paired-choice decision-making algorithms, AHP, GPT-3, or other TLAs (three-letter acronyms).

* Background

James, Andrew, and I worked on Barometer, an effort to use Facebook measurement to promote or demote ads to help defeat Trump in 2020. Even in that effort, we picked what to do next in various different ad-hoc ways. James thought, "Why isn't there a good tool for this?" and convinced us we should try to build it.

I look forward to your feedback and questions. Please don't be cranky.

So... you tried to interfere in an election? I don't get that last part or why that's relevant: maybe you could have said "analyzed potential voter patterns" or some other weasel words. Either way I wish I'd read that before wasting my time trying this (see my other post). Thankfully I can now add you and your "team" to my growing list of activist brogrammers without actual ethics.

Wow, that's pretty harsh.

I don't accept the characterization as a "brogrammer without ethics."

I don't think anything I say would change your opinions, but I'm willing to answer questions if you have any you want answered.

I agree with your assessment that name calling is both presumptuous and unproductive. However, I also think you (perhaps inadvertently at an attempt at comedy) set that tone by assuming that everyone who disagrees with your setup is cranky. There are certainly conversations on here which GPT-3 could predict with striking accuracy, but there are usually good counterpoints interspersed throughout.

In any event, I’m loathe to go down the “election ethics” rabbit hole but I would like to a) reformulate the critique into questions as you’ve asked and b) relate it to your current project.

I find it difficult to believe that you think it’s okay to knowingly sway an election through an apparatus like Facebook. Additionally, Facebook’s ability to do so has already been established by prior research published in Nature:


Since the experiment was simply to increase voting, it was arguably neutral. However, the implication is certainly that an election can be swayed. All you’d have to do is remind one candidate’s base to vote and not the other.

It could be “ethical” to explore nefarious uses of Facebook in order to develop mitigations much in the same way people have talked about gain of function viral research. But, I think what makes your endeavor particularly dubious is the focus on beating Trump.

1) Given FB’s own prior research, what made the answer to your research so valuable that it required pushing generally agreed upon ethnical norms?

2) Given that FB (to my knowledge) does not have an IRB, can you describe what processes you used to make sure you stayed within an ethical framework and to that end, what the framework was?

This is related to your current endeavor because you are asking us to sign in with an organization who also has a, shall we say, “fluid” definition of what “evil” means in their guiding ethical framework which is to “not be”. Ethics aside, it is an organization that already has the ability to triangulate an enormous data from a number sources into an immense amount of power.

So, between your election work and Google’s record, if people are “cranky” it’s probably due to the fact that, as someone astutely pointed out to me once, past behavior is a good predictor of future behavior.

3) So then, how do you plan to establish trust in order for people to use our product?

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact