I was wary of this, as it felt kind of like a horoscope that could apply to anyone.
At the risk of TMI, I left a cult, which was the faith of my family. During that process, I experienced severe spite, grief, and joined religious recovery subreddits to vent, find community, and learn and struggle together with others like me.
Over the years, I wasn't pleased with my tone. I was angry at my previous faith and its leaders for misleading me, and I spoke out about it. Being afraid I was giving too much information about myself on Reddit, and not wanting to get doxxed, I started a new account on Reddit 3 separate times over the years. I ran the analysis on each of the 3, and it captured the thoughts, feelings, and attitudes I had at the time accurately to a point. I'm glad they included the confidence level, because the primary analysis was freakishly spot on, while the lower rungs we're hit and miss.
I'm kind of taken aback. Here, a cold, calculating machine is breaking down the progress I made as an individual. It was a pretty vulnerable experience.
It could totally fake and I could be totally duped, but I don't care enough for it to matter because this rewarding feeling is worth it.
For what it's worth I looked at my reddit account and my hackernews account and they share one personality element in common. My perspective on this is that it's closer to cold reading type stuff - throw out a lot of traits, each of which applies to a lot of people, and then people will naturally find that a lot of what you've said applies to them.
Maybe another way you could test it is give it a day or two until you've forgotten what this website says. Then, write down a short description of yourself at each of the three stages. Finally, check three random accounts and see how much overlap there is between the random accounts and your self-descriptions.
Might also be that people present themselves differently on the two platforms. HN is more about professional discussion while Reddit is more about talking about hobbies.
I love me some tech horoscopes. Fantastic job on this, I can't wait to share this with friends at a bar or similar. It's going to be really entertaining to compare, especially for people who have multiple accounts.
If someone reads the comments first, try this in a private/incognito window please! (Then compare in non-incognito)
There’s a clever product recommendation section, and the test will say “you might own…” and it showed me something I was considering to buy the other day!
It appears there's some rule or other factor that prevents it from working with my name. I thought it might be length, but it does pg just fine. Perhaps because my name is entirely numeric?
It seems to cache names, when I entered mine clicking didn't do anything for the first 10 tries or so, then it was "Attempting to connect to server 1 of 8" for a while, then finally loaded. When I put my name in again it was near instant.
It mostly feels like one of those things you read in horoscopes where it feels right because those things just apply to almost everyone. Sometimes it gives conflicting results, told memy primary theme is energetic but in my secondary theme it said I'm out of energy. That being said "You might be an introvert who still loves to have conversations" that is one of the base characteristics of my personality. So hats of to you
The primary theme was mostly accurate, but the secondary themes were a mix of accurate and completely inaccurate. The needs/desires and recommendation/you might own were completely off. The possible themes were somewhat applicable, but seem like they could be to most anyone.
I don't know if your service is getting hugged or what, but I can see it send off the websocket mouse click events, but it never seems to receive an actionable response. Just regular small packets I assume are keep-alives. Why does it send the screen and client size data back? Fingerprinting? I never see it send either the username, and it mentions the other site when I change it in the UI. What advantage does a websocket offer here over a standard xml http request to a regular API?
Considering some of the comments already in this thread it seems the space considered is significant and at least it's a horoscope that is tailored to the audience, i.e. it's not merely twelve (or so) categories.
It's up to debate I guess if a well tailored horoscope is equivalent to an assessment, but technically both accomplish the same thing. The scary thing is both are indistinguishable from magic from where I stand. And I do put way more faith in the machine as opposed to a stranger who (arguably) knows less facts about me.
Oh no, I gave it a couple of my throwaway accounts, and there was a good amount of overlap in the results. Either it gives everyone similar answers, or my throwaways can be easily linked to each other.
Yea, create some disinformation - but more likely than not, if the AI is any good, the author should be able to differentiate fakes from real connections.
Unless you take extreme precautions, throwaway accounts would be relatively easy to link, by those who run this site. Whether they care, or take the time to do so, is another question. Either way, it can't guarantee anonymity from the most sophisticated actors, it's just a way to keep differing opinions from turning into sustained grudges and personal feuds.
I'm fine with admins being able to link accounts; I'm sure that's very useful for spam prevention. But ideally, I don't want members of the public to be able to link accounts. Maybe you're right; maybe it is hopeless against the most sophisticated actors.
I do indeed own a soda stream. Though I wonder how many do or are close to doing so. Do confirmation biases shift how 'right' it is in total via small specific 'rights' while we overlook 'wrongs'? For example, I do not often fish. To echo mikercampbell.
And is my personality what I type on HN? Are the quite normal people I know really the closet sociopaths they appear to be via their LinkedIn babble?
I might own a hammock for use in place of a tent...that is oddly specific and correct as that is how I camp (when there are appropriate trees available).
> The internet is awesome. You work with tech or maybe have considered it. You probably have a knack for coding, and might love to work for a tech company if you already don't. You are patient and have a passion for the job you do.
I mean this is Hacker News after all (could be guessed just by having an HN username)
It did suggest a Sodastream which was pretty accurate, I already have one.
I also got a Sodastream suggestion which is weird because I don't recall mentioning anything about variety or a desire for soft drinks. I did comment on that notorious Peter Arnell Pepsi pitch deck but that's all that comes to mind.
204 comments
[ 4.1 ms ] story [ 158 ms ] threadAt the risk of TMI, I left a cult, which was the faith of my family. During that process, I experienced severe spite, grief, and joined religious recovery subreddits to vent, find community, and learn and struggle together with others like me.
Over the years, I wasn't pleased with my tone. I was angry at my previous faith and its leaders for misleading me, and I spoke out about it. Being afraid I was giving too much information about myself on Reddit, and not wanting to get doxxed, I started a new account on Reddit 3 separate times over the years. I ran the analysis on each of the 3, and it captured the thoughts, feelings, and attitudes I had at the time accurately to a point. I'm glad they included the confidence level, because the primary analysis was freakishly spot on, while the lower rungs we're hit and miss.
I'm kind of taken aback. Here, a cold, calculating machine is breaking down the progress I made as an individual. It was a pretty vulnerable experience.
It could totally fake and I could be totally duped, but I don't care enough for it to matter because this rewarding feeling is worth it.
Thanks OP. You took me on a journey.
Maybe another way you could test it is give it a day or two until you've forgotten what this website says. Then, write down a short description of yourself at each of the three stages. Finally, check three random accounts and see how much overlap there is between the random accounts and your self-descriptions.
Seriously love projects like this.
Given that I mostly comment on astronomy posts I guess that really is right on the money.
If someone reads the comments first, try this in a private/incognito window please! (Then compare in non-incognito)
There’s a clever product recommendation section, and the test will say “you might own…” and it showed me something I was considering to buy the other day!
This tool needs better UX:
- A loading bar, or some feedback that it is "analyzing data"
- clicking on "Reddit" or "Hacker News" doesn't select the radio buttons
It's up to debate I guess if a well tailored horoscope is equivalent to an assessment, but technically both accomplish the same thing. The scary thing is both are indistinguishable from magic from where I stand. And I do put way more faith in the machine as opposed to a stranger who (arguably) knows less facts about me.
Now the question is: do I get into Internet arguments with people that have a different/opposed primary theme? Could be an interesting follow up.
This seems pretty close.
And is my personality what I type on HN? Are the quite normal people I know really the closet sociopaths they appear to be via their LinkedIn babble?
Well it called me right out.
I mean this is Hacker News after all (could be guessed just by having an HN username)
It did suggest a Sodastream which was pretty accurate, I already have one.