Don't you have a pipeline with girls you're going to cold-call, and a dating life manager that breathes down your neck when you haven't taken at least 10 girls out this month?
hah I love this comment - when I hear mid-30s men calling women "girls" I always cringe. Even worse when they actually call mid-20s women "girls" since they've never dated a women in her 30s..
A dating life manager is a contradiction in itself.
In order to be an expert at dating, it's reasonable to assume that you've been on many dates. That pretty much means you actually suck at dating as the very point of dating is to stop dating (for most).
I was thinking about something similar, but less cringy. More like: hey you haven't messaged your friend in 6 months, maybe you should ask them how they are doing. A priority queue but to help keep friendships alive.
Monica is an excellent "Personal CRM" as people are describing. It strives to achieve exactly the goals GP describes.
Personally, I would just like to see a contacts app that doesn't suck, and actually supports having both "companies" AND "people" as contact without treating them exactly the same. If I want to call some hotel or something, why does it have to be "First name: Holiday Inn" in order to even show up correctly...
MacOS/iOS Contacts supports this, for what it's worth.
You can enter a 'Company Name' and leave fname/lname blank, and it'll give it a different default icon, as well as show it as the company name in lists/sorts.
I remember moaning about twitter reducing the number of characters people sent, and then yo! came out, and it was down to a single bit of data. That made me think 'what would even lower information content messages look like in a social network, perhaps half a bit or even less?'. I figured that a half bit of data would be where a social network sent some sort of hello message automatically with the same frequency the human user sent one manually, so given a message, it'd be 50/50 whether it was sent automatically or manually. Although I got to it theoretically, in the end, I thought it could be pretty useful practically, as a way of sparking conversations again when you haven't spoken to someone in a while, etc.
Maybe combined with ChatGPT we could even make those automatic getting-back-in-touch messages indistinguishable from the manual ones.
Half a bit... yeah. I read a good description of a bit (I think in The Information by James Gleick) as "a yes or no answer to a single unambiguous question. I'm misremembering the quote but that's close enough.
So half a bit would be like hearing "Umm..." in response to that same question?
"Probably" is more than one bit, not less, no? It conveys both uncertainty and an opinion on the answer.
"Is this tree deciduous?"
"Yes" - one bit
"Is this tree evergreen?"
"Probably" - one bit (they don't know), another bit (their guess)
But I'm way off any formal understanding of this, and it can be rigorously defined.
Meta comment: "half a bit" was clearly a joke, and so was my response; now I'm taking your reply at face value and debating it seriously, while admitting that I haven't got anything close to firm enough ground under me to actually debate it :)
I meant based on definition of a bit as "a yes or no answer to a single unambiguous question".
"Umm..." definitely carries some kind of information, but it doesn't actually help answer the question.
Based on the information theoretic definition of entropy we'd need to go from 50:50 to 89% certainty to get half a bit of information, and I'd probably qualify 89% certainty as "probably"
The half-bit you actually send means more to the recipient than the 160 bytes you don't.
I'm more unsettled by the automation part. At least the half-bit received came from someone who consciously thought about you enough to send it. Once you emulate that part, all you have is a MITM initiating conversations between strangers.
> combined with ChatGPT we could even make those automatic getting-back-in-touch messages indistinguishable from the manual ones
extending this bit further - we might as well reply with chatGPT. then it's chatGPT all the way down, who needs human interaction when we tick the boxes of "I called them" and "I replied"?.. :)
In this situation, neither party need be aware that the other party is 'antisocial' in this regard. The way I see it, the following situations apply for a tool that produces communication indistinguishable from the user's own writing (AI)...
"AI" for the following cases would be beneficial:
Social -> Antisocial
Social <- Antisocial
"AI" for the following cases would be net neutral:
Antisocial -> Antisocial
Antisocial <- Antisocial
And finally "AI" would be inapplicable (since neither party would use it) in the following cases:
Social -> Social
Social <- Social
In this drastically simplified model, there aren't any cases where the existence of a sufficiently competent AI would be detrimental to any party involved, while still providing value for those who choose to use it.
To be pedantic, this is not a single bit of data. This is zero bits attached to an event. The event is not discrete however, so I'm not sure what form of information theory would be appropriate to describe it.
Maybe you could also "reduce" the bit in a different way, where it would be 50/50 if a message would actually be sent upon clicking the submit/reply button.
Like if you'd walk on a noisy street, see someone you know, and greet them half-heartedly and wouldn't really mind if they didn't hear you.
Oh man, the thrill of not knowing if this message will actually be sent when I hit the "reply" button!
Lots of attempts over the years at this personal crm problem space but no success to date. I did one on Facebook platform in 2008-09 called Socialfly, our tagline was ‘be twice the friend in half the time’.
Problem is there’s a lot of input and upkeep which limits appeal, the most important social data sources (text & phone) are not accessible via api, its a personal tool that you don’t necessarily want to tell others you use which limits distribution & scale, and even after all that in general it’s hard to scale personal authenticity.
I think the problem needs to be approached from a different angle along the lines of a personal assistant rather that an explicit data management tool. And it likely has to come from those with access to privileged social data sources like Apple or Google or Facebook.
Oh cool, so great to hear from a user after all these years! Some of our users were suuper invested, like keeping up with hundreds of people - far more than we anticipated. I think our top user was a pastor in Australia who kept up with like 700+ people.
We would have continued a while longer but the 08 recession really burnt us. We were raising $$ in the weeks where banks were literally failing. RIP Good Times era. Our iPhone app was just about ready to ship and we were going to start building integrations like email next, but no runway. We had an angel lined up who’d have gotten us about 4-6mo of runway but I didn’t want to burn that relationship in such an uncertain environment, we were all pretty down about prospects then.
Fun fact tho: the app continued to work for years after. FB took a long time to deprecate API and our code was pretty solid.
Numbers yes, when you last contacted no. Core value prop of personal CRM is ease of keeping up w/ people, which means smarts of reprioritizing people based on several factors sort of like a personal assistant would, which means system should know when you last spoke/texted/emailed/etc. Ideally it'd even have ease of access to transcripts (chat/email logs) so you could jog your memory.
The goal is to avoid just a dumb timer saying 'its been 4 weeks, why don't you chat with X'. That's how Socialfly worked, and it was pretty high friction in practical use.
There have been several iterations of this sort of idea — a personal CRM — and the main issue I have is they should automatically scan my emails, texts and chats to figure out my friends and suggest that I reach out.
> should automatically scan my emails, texts and chats
Great, easy to integrate one service with the APIs of a bunch of others!
> Also, want it to be 100% local and privacy first.
...but in combination? Never going to happen.
Ask yourself: who's signing up for the API keys to enable the client-side service to talk to all these services? Is it the end-user, or is it this software's developer?
If it's the software's developer, then they're effectively leaking all these API keys by embedding them into the software itself — where not just the end users, but anyone else could come along and reuse these keys for anything they like. The service providers will find this out, and block these keys. (No, you can't avoid this by proxying requests to some gateway, operated by the service-provider, that holds the API keys. Then you lose the "local/private" aspect.)
If it's each end-user, then the aggregate traffic from all the instances of this app running at once, will look exactly like a bot that's trying to evade API rate-limits using a "residential proxy cluster" like https://www.zyte.com/smart-proxy-manager/... and so the services will block these keys.
---
Mind you, in theory, you could do this on the OS level, using OS accessibility APIs to effectively "read" the messages off the screen. But 1. is there any third-party ISV — who isn't a certified accessibility-software provider — who you'd trust enough to allow their software the ability to constantly "read" everything on your screen? That includes your passwords, you know! And also, 2., the messages need to be on the screen for accessibility software to read them. An accessibility-API-driven CRM can't load your chat history unless you also grant it the ability to literally take your mouse and scroll through it for you.
Or, alternately, coming at this from the perspective of the Operating System vendor themselves, you could do this "in" the OS, by forcing emails/text/chat message handling to go through system APIs that can see these as special document types, and so do things with them. IIRC there was at least one pre-iPhone mobile OS that did this (BlackBerry OS, maybe?), enabling all of these types of messaging-app traffic to be muxed together into a single first-party app that did indeed manage all conversations with your contacts in a multi-channel way.
That presumes the client stores anything locally + in plaintext. Clients like Facebook Messenger, Slack, and Discord have no local storage other than (opaque, fragmentary, useless) cache databases, because they're just "webapps in a can"; while many of the more modern native chat clients like WhatsApp, Signal, Telegram, etc keep all their synced state encrypted-at-rest on disk.
Also, that presumes you're on desktop. On mobile, you just can't poke into another app's sandbox/container that way (unless the target app hasn't explicitly granted made some part of its container-filesystem externally accessible; and there's no reason these apps would.) And many people only have a mobile device.
If you're okay with a partial solution, a far easier one would be scraping the data out of the few clients that have web-app versions through browser extensions that read the state out of the page they're running in.
> I was thinking about something similar, but less cringy.
Well, if you are worried about it being cringy, you probably shouldn't watch the 2022 Salesforce Dreamforce (their annual convention). It's cultish and shows the unbelievably arrogant confidence they have in their own importance - along with the most inexplicably childish moments for a professional conference I've ever seen. (Let's put foam rabbit ears on our "co-CEOs" to entertain the "trailblazers" as we gather around a fake wooden stage imitating the outdoors with faux trees, won't that be funny?)
This combined with the only useful feature FB had for years (people's birthday).
Integrate a scraping and keyword search for public facing social media postings to alert to people in distress, perhaps. Or just a general "you should check on X" feature. Not with automated messaging, that's too much.
> hey you haven't messaged your friend in 6 months, maybe you should ask them how they are doing. A priority queue but to help keep friendships alive.
This is what social media and messaging apps should innovate on.
But, without VC money, without ads, without feeds drive by algorithms that optimize for engagement. Without dark patterns and antifeatures that make our lives worse. Just a non-profit that churns out open source code and is supported by monthly recurring donations from a % of the userbase.
I want one for remembering stuff about people for personal and business networking.
For a while, I was going to a weekly event where I'd meet people, but I'd never remember their name next time because I'm terrible with names.
So I started jotting down notes after it was over. Like, today I met person X who is really into this one hobby and just moved here from such and such city. And person Y who is looking for a new apartment. And Z who came with Y.
Then before going next time, I'd review the list, and if I saw X, I knew their name and could ask how they're adjusting after their move. Or I could ask Y how the apartment search is going.
At first I thought if people saw this list they might think it was a little weird. And maybe, but I'm OK with it since it's a way of making an effort. As long as it's genuine and your motivations are good, people like that you remembered stuff about them. (I'm not doing it to impress people, etc.)
Anyway, I didn't have a good way to organize it. I stuck it all in a document, which didn't work great.
I've been building FriendApp the past year. The core feature on iPhone right now is to just group people into lists, and preserve messaging integration with whatsapp, text etc. Trying to make it super simple. Also has ability to sync for recent contacts list. Would love feedback, and thoughts on the next features to iterate on.
I signed up for https://infrequent.app/ when it was a Show HN just to try it out and it actually worked (probably about as well as a "non-busy" TODO in my calendar app with an alert). Enough to remind me to call 1 or 2 people that I like to talk to but often leave for too long, anyway.
I've thought about similar problems like rewatching your favorite movie, and concluded that treating it as a priority queue or trying to watch it every _n_ days is the wrong approach; you want to instead be reminded when you've forgotten about your friend Alex or what _The Leopard_ is about, so you can then ping them & catch up or rewatch that movie with a near-fresh mind. (If you ping Alex while you remember them, then maybe there's another actually-forgotten friend you could've pinged instead who would be more valuable to get back in touch with.)
It is, in other words, the exact opposite of 'spaced repetition', where you want to review right before forgetting to strengthen the memory the most; in this use-case, 'anti-spaced repetition', you want to review only after you've forgotten it. (https://www.gwern.net/Statistical-notes#program-for-non-spac...)
This is where I'd like to go with FriendApp. Would love if you wanted to check out where we are right now with our iPhone app. https://www.friendapp.com/
Thing is it has nothing to do with how some exec feel about the product.
Corporations don't like stuff like because they don't want to deal with clients who don't understand the joke being playing with their brand. OP isn't going to be the one who has to do damage control when this blows up and some percentage of people start to to believe it's actual Salesforce behind this product.
Might be a good opportunity here to integrate events, restaurant bookings, and other paid "experiences" for some upside.
Does this integrate directly with the dating services/apps themselves? Seems like a lot of work to plumb in and keep up-to-date the status and latest information on each profile.
Also, if this is a legit service you should be extremely careful how you store all this data. Exposing a bunch of dating profile data via a security breach will land you into hotter water than anything Marc Benioff could throw at you :)
When I was dating, I kept track of all this manually, in a notes app. I wouldn't mind inputting the same data here, manually, since this has the added benefit of organizing it for me in a way I couldn't do with mere notes.
Then for dates in the funnel who told you things like "too busy", "not ready for dating" etc. you could schedule automated messages for things like reminding them to circle back when they're ready, checking if they're available to get on a call to show them new features you've added (as a person), etc.
I assume this is humor, since a lot of people will use these responses to say "No" without having to come out an say it (the "Nonconfrontational No", if you will - "I don't want to date you" invites a direct conflict, but "too busy" can't really be argued with).
So, building on your humor, I'd like to suggest a feature/service we can sell to the victcoughother person. When they respond with "too busy" they can also set a flag in the system indicating whether they'd actually like to _not_ hear from the person operating the funnel. We get paid, the vict^H^H^H^Hdate doesn't get hassled in the future, and the person running the dating funnel is none the wiser because our software never notifies them again.
> Hi (Name), are you getting the best experiences on your dates? We've been chatting for a while, and I wanted to share some news with you before I update everyone else. I've found some new exciting dating spots and you're going to love them! (Insert sales pitch) I'd love the chance to chat about my new dating spots. Shall we arrange a meeting to talk you through possible restaurants, bowling alleys or what movies we can watch together? Let's book a call today and get things started.
Can someone chime in why this hasn't worked in the past?
It seems a personal/portable CRM could be highly useful.
The second aspect is the vertical aspect (professional or "personal")
One thing I keep hearing from people that have sold companies, exited etc. is that they are having a hard time operating within their networks without the CRM.
Yes. Having been in this space for a bit, there are a few reasons why it hasn't worked (yet):
1. The problem of deduplicating contacts is tough (but solvable). If you don't solve it well, then the utility of personal CRM goes wayyyyy down and you're getting notifications about the same people with different email / WhatsApp / Instagram addresses and that gets annoying.
2. People haven't shown a big willingness to pay very much for this (so far), despite everyone saying they want it. So we're left with the open source solutions that don't solve the problems very well.
3. People are SUPER concerned about privacy. See a comment above about someone who wants the system to automatically scan email, etc. and extract contacts, but that it must be 100% local and privacy-centric. You can't have both of those - to intelligently extract contacts without duplicates (and figure out that @mybestie on Instagram == mybestie@gmail.com == my.c.bestie@corporate.com) you need a big database of who's who to match against.
Everything is solvable I think - we're not talking cold fusion here. But it's tougher than it might seem on the surface.
Could you expand on what options exist currently? You say you are in the space, does that mean you have been keeping an eye on it, that you are working on something, or ?
I'm in the space in that I co-founded a company called Connect The Dots (ctd.ai) that certainly has the potential to solve this problem, but we're not focused on the PRM space right now (see point above re: unwillingness to pay). Perhaps down the road.
One problem I've had adopting such a system is that you want the system to help manage some load of work, e.g. maintaining your personal relationships. But using the product is itself work: you need to set it up, remember to use it, keep it up to date, etc. So, just as you've procrastinated keeping up with old friends, you procrastinate using the tool. And, of course, the tool doesn't reduce the workload of keeping up with friends. In fact, done successfully, your work has increased as you need to talk to your friends more often. So you're basically adding work (using the tool) to do work you didn't have the energy for in the first place (talking to your friends).
Everyone loves the idea of talking to their friends, conceptually. But, given ample time and opportunity, many choose not to keep up with friends. It's almost as if people only "want to want" to keep up with friends, rather than actually want it.
Yes yes absolutely 100x. That's another problem with the existing tools - they're a lot of work! You can say "we'll solve the duplicate contacts problem by ignoring it and letting using de-dupe themselves" welllll that's a lot of work!
And then there's the empty-database problem. I installed Monica a while back and the first thing it prompted me to do was.....input all my contacts. No Facebook import, no LinkedIn import, etc. Are you kidding me? I'm not rebuilding my friend graph from scratch. So I ditched the tool immediately (note: I haven't looked at it in years so it might have fixed some of those issues in the meantime).
And absolutely yes - there's the actual work of being in touch with people that you can't (or shouldn't, in a non-dystopian world) automate away. I'm sure someone will figure out how to rig up GPT-3 to do that but ughhhh...shiver please don't.
I believe a plugin (e.g. in Obsidian, Notion, Superhuman) could work better than Monica. In fact I think Superhuman has the bets oppportunity to create this product.
CRMs work at scale by selling to executives who are not the end users but are in control of budget. They then have their sales managers enforce data fidelity amongst the sales team who, to keep their jobs, are incentivized to make sure data in the CRM are up to date, or whatever the closest approximation to that is.
So for a personal CRM to work, you'd need to sort out, at minimum, the monetization piece and the data fidelity piece. If you open source it, you still need to make sure people keep the data up to date and that's actually pretty hard.
Executives also want to take the rolodex with them, hence they have an incentive to keep a personal CRM so that they keep their contacts after they leave.
Superhuman sold to executives for $30/month - which is a relevant (high) price point.
I'm not entirely sure what the first point has to do with the second - can you help me understand what Superhuman has to do with anything? It's just a gmail client at the end of the day, isn't it?
My understanding is people who want to keep their contacts have 1) a lot of them in the actual inbox 2) have their Linkedin to fall back on 3) can always fall back to exporting all their contacts when they leave.
Apropos of nothing, $30/month/user is roughly the base price of Salesforce, and they have 3 more tiers significantly above that. High for a consumer, but pretty cheap for professional tools, and they don't sell licenses one at a time.
I think it simply would need a lot of discipline to keep up to date and without a manager nagging or bonuses being dependent upon updating it, most people lack that discipline.
> Can someone chime in why this hasn't worked in the past?
Because spreadsheets and notes are free and good enough.
I used to be a full time shitty cheap car flipper. I used to be single. They're about the same level of communication work. The ROI of a proper CRM isn't there because in both cases your customer is just not that serious about the interaction when they're in the part of the funnel a CRM helps you with and even then it doesn't really help you with the bulk of the customers, it helps you with the long tail. When you're a one man shop squeezing out an extra 1% or whatever a CRM gets you isn't worth the time vs focusing on other areas. (These days I work in a client facing role with a proper ticketing system and integrated CRM so I do have something to compare to.)
I've been working on FriendApp the past year. I think it hasn't worked in the past because the average person doesn't realize that they could really benefit from this solution. It hasn't necessarily felt like a burning problem. I think it has to add some value beyond just classifying and adding additional metadata to contacts.
My goal is to develop features that also facilitate sharing of activities, upcoming events that I'm attending that I want to share with people I know. Things I want to do, and making that visible to select groups of people.
Dunno if this was intended as a joke, but I find the whole presentation of the site to be hilarious. Well done!
As others have said, if you were able to find if Hinge, Bumble, Tinder, Grindr, etc... have open APIs that you could connect to your app? Same goes for things like Eventbrite or Meetup.
"I met so-and-so through (dating app) on (date/time) and we hit it off well. We then met up again via (event app) and continued meeting up regularly, and I was able to keep notes about so-and-so on the app - their favorite bars, foods, music, etc... as they revealed it to me, and the app was able to provide me with info as to when I was last in touch with them, when there's availability at their favorite restaurant, a music act at their favorite bar, etc..."
As much as I love this idea, part of me read this and thought "Great, one more reason for women who aren't interested in me to feign even more interest, because big event xyz is coming up and it's 'revealed' to them I can afford it so it's 'revealed' to me they want to go."
I laughed when I read this but then I thought more and... wouldn't that be fun? Like, if it's their hobby and they're good at it, why would the date be bad?
Gardeners like their plants to grow big and healthy. Fishkeepers obsess over the health of their tanks. Musicians want to play more challenging music with better bandmates for bigger audiences on higher profile stages.
So if you were on the receiving end of a date orchestrated by a dating hobbyist, it'd be fun, probably unusual, definitely well organized. There are worse things to experience while dating. And, just maybe, you make a connection and form a relationship. But even the base case isn't _bad_ per se.
I have never heard that definition of "hobbyist" before. I don't see how it's even remotely related to what I wrote. But whatever. Get some rest, maybe tomorrow will be a better day for you.
Dating has a large emotional aspect. Using heavyweight management software for this kind of thing is ridiculous. If you need CRM to remind yourself of the emotional impact you had on a date, then some part of this process has gone horribly wrong.
Honestly, this sounds like the gulf between men’s experience dating and women’s, and neurotypical vs neurodivergent. Some ND people like helpers like this; and, women get so many suitors that a system to help them may be actually beneficial.
Wow, managing one's dating life with a crm app has got to be the most distopian thing I've seen in a while.
I probably need to point out that I'm not dissing the app. If it succeeds, the creators are obviously filling a need in the market, but confirmation of such a need would indicate dating as I define it is horribly broken
Socializing as a whole is broken, triggered by the disappearance of third places, the dissolution of communities at the altar of the nuclear dual-income family, live-at-work jobs and "hustle culture" (Veblen entrepreneurship), and the unlimited streaming doom-scroll.
With nowhere to go but work and home, good luck finding anyone to date or befriend via a nonexistent organic social network.
While what you are saying is very true, dating so many people at once, plus having dated so many in the past that you need an app to keep track of them all is a whole other level of broken
The issue is that whereas in the past you could use "knowing somebody as an acquaintance and/or by association with an acquaintance" as the first stage of your "should I date somebody" funnel, you now get "here's a list of all the people in your who have signed up for a dating service, the marketing blurb they wrote about themselves, and some broad demographic filters; Have at it" as your initial filter.
Recently an article described an extreme case, a woman in her thirties wasting 10 years "swiping" and dating "at scale". She'd check over a 100 profiles every single day for 10 years straight without even an attempt at forming a relationship.
Pure addiction, like a slot machine. The original purpose all but forgotten.
A sizeable group of us don’t have a meaningful second place either anymore. I work remote and there is really no social aspect to it in any meaningful sense, just pointless meetings and an endless Jira backlog. My home and my work are one and the same.
In the last three years the closest I came to a viable third place was VRChat, it was the best social experience I’ve ever had. I don’t think I need to say much more about that, it speaks for itself.
242 comments
[ 4.3 ms ] story [ 305 ms ] thread> P.S. Marc Benioff, please don't sue us. Remember that imitation is the sincerest form of flattery.
I never realized that my dream job would be so people-oriented. Drop me a line if you're hiring...
In order to be an expert at dating, it's reasonable to assume that you've been on many dates. That pretty much means you actually suck at dating as the very point of dating is to stop dating (for most).
Personally, I would just like to see a contacts app that doesn't suck, and actually supports having both "companies" AND "people" as contact without treating them exactly the same. If I want to call some hotel or something, why does it have to be "First name: Holiday Inn" in order to even show up correctly...
* https://keepmyfriends.com/ * https://getdex.com/
I haven't used any personally!
Maybe combined with ChatGPT we could even make those automatic getting-back-in-touch messages indistinguishable from the manual ones.
I’ve never heard that idea before and I love it.
So half a bit would be like hearing "Umm..." in response to that same question?
"Is this tree deciduous?" "Yes" - one bit
"Is this tree evergreen?" "Probably" - one bit (they don't know), another bit (their guess)
But I'm way off any formal understanding of this, and it can be rigorously defined.
Meta comment: "half a bit" was clearly a joke, and so was my response; now I'm taking your reply at face value and debating it seriously, while admitting that I haven't got anything close to firm enough ground under me to actually debate it :)
"Umm..." definitely carries some kind of information, but it doesn't actually help answer the question.
Based on the information theoretic definition of entropy we'd need to go from 50:50 to 89% certainty to get half a bit of information, and I'd probably qualify 89% certainty as "probably"
To try to be a little constructive, is a relationship really worth anything if a half-bit is all one can spare for it?
I'm more unsettled by the automation part. At least the half-bit received came from someone who consciously thought about you enough to send it. Once you emulate that part, all you have is a MITM initiating conversations between strangers.
extending this bit further - we might as well reply with chatGPT. then it's chatGPT all the way down, who needs human interaction when we tick the boxes of "I called them" and "I replied"?.. :)
"AI" for the following cases would be beneficial:
Social -> Antisocial
Social <- Antisocial
"AI" for the following cases would be net neutral:
Antisocial -> Antisocial
Antisocial <- Antisocial
And finally "AI" would be inapplicable (since neither party would use it) in the following cases:
Social -> Social
Social <- Social
In this drastically simplified model, there aren't any cases where the existence of a sufficiently competent AI would be detrimental to any party involved, while still providing value for those who choose to use it.
To be pedantic, this is not a single bit of data. This is zero bits attached to an event. The event is not discrete however, so I'm not sure what form of information theory would be appropriate to describe it.
Like if you'd walk on a noisy street, see someone you know, and greet them half-heartedly and wouldn't really mind if they didn't hear you.
Oh man, the thrill of not knowing if this message will actually be sent when I hit the "reply" button!
Problem is there’s a lot of input and upkeep which limits appeal, the most important social data sources (text & phone) are not accessible via api, its a personal tool that you don’t necessarily want to tell others you use which limits distribution & scale, and even after all that in general it’s hard to scale personal authenticity.
I think the problem needs to be approached from a different angle along the lines of a personal assistant rather that an explicit data management tool. And it likely has to come from those with access to privileged social data sources like Apple or Google or Facebook.
We would have continued a while longer but the 08 recession really burnt us. We were raising $$ in the weeks where banks were literally failing. RIP Good Times era. Our iPhone app was just about ready to ship and we were going to start building integrations like email next, but no runway. We had an angel lined up who’d have gotten us about 4-6mo of runway but I didn’t want to burn that relationship in such an uncertain environment, we were all pretty down about prospects then.
Fun fact tho: the app continued to work for years after. FB took a long time to deprecate API and our code was pretty solid.
The goal is to avoid just a dumb timer saying 'its been 4 weeks, why don't you chat with X'. That's how Socialfly worked, and it was pretty high friction in practical use.
Also, want it to be 100% local and privacy first.
But I also want an AI personal assistant to know what to write to everyone, what we last did, what would they be interested in, and so on.
https://github.com/monicahq/monica
There’s also the problem of entering unflattering details, and subsequently leaking them out.
The search continues.
Great, easy to integrate one service with the APIs of a bunch of others!
> Also, want it to be 100% local and privacy first.
...but in combination? Never going to happen.
Ask yourself: who's signing up for the API keys to enable the client-side service to talk to all these services? Is it the end-user, or is it this software's developer?
If it's the software's developer, then they're effectively leaking all these API keys by embedding them into the software itself — where not just the end users, but anyone else could come along and reuse these keys for anything they like. The service providers will find this out, and block these keys. (No, you can't avoid this by proxying requests to some gateway, operated by the service-provider, that holds the API keys. Then you lose the "local/private" aspect.)
If it's each end-user, then the aggregate traffic from all the instances of this app running at once, will look exactly like a bot that's trying to evade API rate-limits using a "residential proxy cluster" like https://www.zyte.com/smart-proxy-manager/... and so the services will block these keys.
---
Mind you, in theory, you could do this on the OS level, using OS accessibility APIs to effectively "read" the messages off the screen. But 1. is there any third-party ISV — who isn't a certified accessibility-software provider — who you'd trust enough to allow their software the ability to constantly "read" everything on your screen? That includes your passwords, you know! And also, 2., the messages need to be on the screen for accessibility software to read them. An accessibility-API-driven CRM can't load your chat history unless you also grant it the ability to literally take your mouse and scroll through it for you.
Or, alternately, coming at this from the perspective of the Operating System vendor themselves, you could do this "in" the OS, by forcing emails/text/chat message handling to go through system APIs that can see these as special document types, and so do things with them. IIRC there was at least one pre-iPhone mobile OS that did this (BlackBerry OS, maybe?), enabling all of these types of messaging-app traffic to be muxed together into a single first-party app that did indeed manage all conversations with your contacts in a multi-channel way.
Look into it, you might be surprised.
You were very active with this friend for n years and it dropped off. Drop them a line.
Also, that presumes you're on desktop. On mobile, you just can't poke into another app's sandbox/container that way (unless the target app hasn't explicitly granted made some part of its container-filesystem externally accessible; and there's no reason these apps would.) And many people only have a mobile device.
If you're okay with a partial solution, a far easier one would be scraping the data out of the few clients that have web-app versions through browser extensions that read the state out of the page they're running in.
Well, if you are worried about it being cringy, you probably shouldn't watch the 2022 Salesforce Dreamforce (their annual convention). It's cultish and shows the unbelievably arrogant confidence they have in their own importance - along with the most inexplicably childish moments for a professional conference I've ever seen. (Let's put foam rabbit ears on our "co-CEOs" to entertain the "trailblazers" as we gather around a fake wooden stage imitating the outdoors with faux trees, won't that be funny?)
Integrate a scraping and keyword search for public facing social media postings to alert to people in distress, perhaps. Or just a general "you should check on X" feature. Not with automated messaging, that's too much.
This is what social media and messaging apps should innovate on.
But, without VC money, without ads, without feeds drive by algorithms that optimize for engagement. Without dark patterns and antifeatures that make our lives worse. Just a non-profit that churns out open source code and is supported by monthly recurring donations from a % of the userbase.
For a while, I was going to a weekly event where I'd meet people, but I'd never remember their name next time because I'm terrible with names.
So I started jotting down notes after it was over. Like, today I met person X who is really into this one hobby and just moved here from such and such city. And person Y who is looking for a new apartment. And Z who came with Y.
Then before going next time, I'd review the list, and if I saw X, I knew their name and could ask how they're adjusting after their move. Or I could ask Y how the apartment search is going.
At first I thought if people saw this list they might think it was a little weird. And maybe, but I'm OK with it since it's a way of making an effort. As long as it's genuine and your motivations are good, people like that you remembered stuff about them. (I'm not doing it to impress people, etc.)
Anyway, I didn't have a good way to organize it. I stuck it all in a document, which didn't work great.
https://www.friendapp.com/
It is, in other words, the exact opposite of 'spaced repetition', where you want to review right before forgetting to strengthen the memory the most; in this use-case, 'anti-spaced repetition', you want to review only after you've forgotten it. (https://www.gwern.net/Statistical-notes#program-for-non-spac...)
- MonicaHQ (open-source)
- Dex (free version seems good enough for most people, i'm trying this one out rn) getdex.com/
- Custom system in Airtable - https://news.ycombinator.com/item?id=30329475
- Infrequent app - blog post talks more about it, linked in comments - https://news.ycombinator.com/item?id=33308084
Corporations don't like stuff like because they don't want to deal with clients who don't understand the joke being playing with their brand. OP isn't going to be the one who has to do damage control when this blows up and some percentage of people start to to believe it's actual Salesforce behind this product.
If you'd please review https://news.ycombinator.com/newsguidelines.html and stick to the rules when posting here, we'd appreciate it.
Does this integrate directly with the dating services/apps themselves? Seems like a lot of work to plumb in and keep up-to-date the status and latest information on each profile.
Also, if this is a legit service you should be extremely careful how you store all this data. Exposing a bunch of dating profile data via a security breach will land you into hotter water than anything Marc Benioff could throw at you :)
Then for dates in the funnel who told you things like "too busy", "not ready for dating" etc. you could schedule automated messages for things like reminding them to circle back when they're ready, checking if they're available to get on a call to show them new features you've added (as a person), etc.
So, building on your humor, I'd like to suggest a feature/service we can sell to the victcoughother person. When they respond with "too busy" they can also set a flag in the system indicating whether they'd actually like to _not_ hear from the person operating the funnel. We get paid, the vict^H^H^H^Hdate doesn't get hassled in the future, and the person running the dating funnel is none the wiser because our software never notifies them again.
I think I'm ready for an MBA now :)
> Hi (Name), are you getting the best experiences on your dates? We've been chatting for a while, and I wanted to share some news with you before I update everyone else. I've found some new exciting dating spots and you're going to love them! (Insert sales pitch) I'd love the chance to chat about my new dating spots. Shall we arrange a meeting to talk you through possible restaurants, bowling alleys or what movies we can watch together? Let's book a call today and get things started.
Source: https://www.flowrite.com/blog/sales-pitch-email , slightly modified.
Wow, you even have it running on NA14?
It seems a personal/portable CRM could be highly useful. The second aspect is the vertical aspect (professional or "personal")
One thing I keep hearing from people that have sold companies, exited etc. is that they are having a hard time operating within their networks without the CRM.
1. The problem of deduplicating contacts is tough (but solvable). If you don't solve it well, then the utility of personal CRM goes wayyyyy down and you're getting notifications about the same people with different email / WhatsApp / Instagram addresses and that gets annoying.
2. People haven't shown a big willingness to pay very much for this (so far), despite everyone saying they want it. So we're left with the open source solutions that don't solve the problems very well.
3. People are SUPER concerned about privacy. See a comment above about someone who wants the system to automatically scan email, etc. and extract contacts, but that it must be 100% local and privacy-centric. You can't have both of those - to intelligently extract contacts without duplicates (and figure out that @mybestie on Instagram == mybestie@gmail.com == my.c.bestie@corporate.com) you need a big database of who's who to match against.
Everything is solvable I think - we're not talking cold fusion here. But it's tougher than it might seem on the surface.
The two commercial tools that come to mind are:
* Dex (https://getdex.com/)
* Clay (https://clay.earth/)
Although neither have really nailed it (yet). Clay is more like a Superhuman-for-Contacts, whereas Dex is really going hard at being a true PRM.
The big open source tool out there is:
* Monica (https://www.monicahq.com/)
Everyone loves the idea of talking to their friends, conceptually. But, given ample time and opportunity, many choose not to keep up with friends. It's almost as if people only "want to want" to keep up with friends, rather than actually want it.
And then there's the empty-database problem. I installed Monica a while back and the first thing it prompted me to do was.....input all my contacts. No Facebook import, no LinkedIn import, etc. Are you kidding me? I'm not rebuilding my friend graph from scratch. So I ditched the tool immediately (note: I haven't looked at it in years so it might have fixed some of those issues in the meantime).
And absolutely yes - there's the actual work of being in touch with people that you can't (or shouldn't, in a non-dystopian world) automate away. I'm sure someone will figure out how to rig up GPT-3 to do that but ughhhh...shiver please don't.
So for a personal CRM to work, you'd need to sort out, at minimum, the monetization piece and the data fidelity piece. If you open source it, you still need to make sure people keep the data up to date and that's actually pretty hard.
Superhuman sold to executives for $30/month - which is a relevant (high) price point.
My understanding is people who want to keep their contacts have 1) a lot of them in the actual inbox 2) have their Linkedin to fall back on 3) can always fall back to exporting all their contacts when they leave.
Apropos of nothing, $30/month/user is roughly the base price of Salesforce, and they have 3 more tiers significantly above that. High for a consumer, but pretty cheap for professional tools, and they don't sell licenses one at a time.
Because spreadsheets and notes are free and good enough.
I used to be a full time shitty cheap car flipper. I used to be single. They're about the same level of communication work. The ROI of a proper CRM isn't there because in both cases your customer is just not that serious about the interaction when they're in the part of the funnel a CRM helps you with and even then it doesn't really help you with the bulk of the customers, it helps you with the long tail. When you're a one man shop squeezing out an extra 1% or whatever a CRM gets you isn't worth the time vs focusing on other areas. (These days I work in a client facing role with a proper ticketing system and integrated CRM so I do have something to compare to.)
My goal is to develop features that also facilitate sharing of activities, upcoming events that I'm attending that I want to share with people I know. Things I want to do, and making that visible to select groups of people.
Hope you'll check out where we're heading https://www.friendapp.com/
As others have said, if you were able to find if Hinge, Bumble, Tinder, Grindr, etc... have open APIs that you could connect to your app? Same goes for things like Eventbrite or Meetup.
"I met so-and-so through (dating app) on (date/time) and we hit it off well. We then met up again via (event app) and continued meeting up regularly, and I was able to keep notes about so-and-so on the app - their favorite bars, foods, music, etc... as they revealed it to me, and the app was able to provide me with info as to when I was last in touch with them, when there's availability at their favorite restaurant, a music act at their favorite bar, etc..."
I'd love to have an app like that in my life.
SMH
Gardeners like their plants to grow big and healthy. Fishkeepers obsess over the health of their tanks. Musicians want to play more challenging music with better bandmates for bigger audiences on higher profile stages.
So if you were on the receiving end of a date orchestrated by a dating hobbyist, it'd be fun, probably unusual, definitely well organized. There are worse things to experience while dating. And, just maybe, you make a connection and form a relationship. But even the base case isn't _bad_ per se.
> dating hobbyist
Also, your use of hobbyist is so close to being on par (https://www.urbandictionary.com/define.php?term=hobbyist).
I have never heard that definition of "hobbyist" before. I don't see how it's even remotely related to what I wrote. But whatever. Get some rest, maybe tomorrow will be a better day for you.
Also, journaling is a powerful tool.
> Too many requests, please try again after an hour
I’m really out of touch
I probably need to point out that I'm not dissing the app. If it succeeds, the creators are obviously filling a need in the market, but confirmation of such a need would indicate dating as I define it is horribly broken
Socializing as a whole is broken, triggered by the disappearance of third places, the dissolution of communities at the altar of the nuclear dual-income family, live-at-work jobs and "hustle culture" (Veblen entrepreneurship), and the unlimited streaming doom-scroll.
With nowhere to go but work and home, good luck finding anyone to date or befriend via a nonexistent organic social network.
While what you are saying is very true, dating so many people at once, plus having dated so many in the past that you need an app to keep track of them all is a whole other level of broken
Pure addiction, like a slot machine. The original purpose all but forgotten.
In the last three years the closest I came to a viable third place was VRChat, it was the best social experience I’ve ever had. I don’t think I need to say much more about that, it speaks for itself.
I agree if it becomes a social norm, but it could also be really helpful for people with certain disabilities or even dementia