Hacker News new | past | comments | ask | show | jobs | submit login
Ask HN: Help me generate an idea for my next project
34 points by chocolatejb on Jan 21, 2018 | hide | past | favorite | 28 comments
Ask HN: Generating ideas for my next project.

My previous one was [CryptoMilli](https://www.cryptomilli.com). In the back of my mind, I want to work on a single page app with the ethereum blockchain as the backend.

With the amount of energy it generally takes, I want to at least work on something worthwhile. That's why I'm giving myself a week or two to evaluate a bunch of ideas, choose one and ditch the others. (Dating a bunch to find out which one I could fall in love with, at least until it's done).

Any pointers, suggestions will be highly appreciated.




Make a digitally verified lab book

It would be kind of a wiki but each update/edit is externally verified.

In science, each lab has several lab books where everybody documents everything they did. In the rare event of a patent discussion between two labs or unis the first entry in the lab book decides who gets the patent (it seems to be way more complicated than that by now). Usually these lab books are regulary counter signed by the supervisor to 'prevent' tampering but it's not a particularly secure system. Having the blockchain verify each entry's date and originality could be fun. (but I wonder how useful it would be in the real world - would patent lawyers take this up or just ignore it and say they're happy with the old system?)


In my lab experience, electronic lab notebooks were incredibly bloated, slow, and buggy to the point of being unusable. I’m most familiar with what was in place at Merck about 8 years ago: ELN (Electronic Lab Notebook). I agree with comments that the countersignature is both lax and absolutely necessarily from a records perspective. Most importantly, building a new lab notebook system would require importing data from whatever it is you’d be attempting to displace. Lab notebooks have significant roles in IP disputes, and it would be critical that the contents and timestamps are preserved.

To underscore how terrible ELN was at our site, many managers instructed their reports to keep traditional paper notebooks, photograph each page, and upload to an ELN entry. Countersignatures were then performed on the ELN platform. This was at a relatively small Merck site, and maybe the issue wasn’t as pervasive elsewhere.


Just wanted to say that I think this is a interesting idea, and that labs often have 'signing parties' where they do the counter signing.


Forgot something - most wetlab laboratory lab-books are physical copies with handwritten content, mostly so you can quickly glue in photos and take notes in spaces where PCs are not allowed, this could also hamper uptake of any PC-based lab-book.


Build an ethereum contract (or verify and use an existing one) that refunds all money received unless a 'goal' is met, like a 'guaranteed money back if useless' GoFundMe.

Allow people to submit ideas that you're willing to vet completion for (this gets tricky/arbitration-y, encourage splitting this up into phases/steps rather than a giant lump sum) like an 'only pay if you get it' Kickstarter/Upwork.

With a twist: find a way to use all the pending deposits for proof of stake and return most of what is earned to those who have committed to potential projects. Consider making your own transaction fee equal to the ethereum transaction fee for all your transactions.

I may spam a couple more as separate comments...


First find a problem, then think of the obvious solution to that problem, then solve the underlying problem to that solution.

You can find some good problems from this essay.

https://medium.com/black-n-white/the-problem-with-problems-4...


Thanks for sharing ThomPete I read the article. I will go through the HN thread link you pointed out in a bit.


If you’ve decided to use Ethereum, you should find a target group that really needs the immutability and verified consensus. For the average user, Ethereum is both very expensive and very inconvenient.

So maybe look for industries where they’re currently moving a lot of paper around to achieve consensus between market players and/or regulators?


Thanks a lot for the suggestion Pavlov. I will let this float in my head and search/read more about some use-cases.


Design a polling system for media houses to use when they conduct opinion polls for a particular topic. Making it decentralized and blockchain backed will ensure truth of data and make the results tamper proof.


I got a pet project in concept stage :) Very experimental though, no guaranties at all it will work:

A mix of link aggregation and twitter (taking the best of both worlds) and than adding hierarchy by ratings from the link curators. This hopefully means avoiding suggestion engine clickbait, and democratic mainstream kitten pictures. It's always the 1 to 1 trust of subscriber and curator. This also goes away from timeliness and chronology.

Further experimental features:

- adjust your subscriptions by dampening or raising certain curators weighting

- meta subscriptions: Subscribe everything someone else subscribed to (with dampening when there are too many nested hops). Should take viral to an extreme (for good or bad).

- see each post only once (flip or scroll to the next page and it becomes permanently archived for you)

- nested folders and subscription based on folders (splitting personalities and allowing to split topics. In combination with meta-subscriptions one could subscribe to accounts of resort editors to make a "magazine")

If you are interested, I wrote a few pages about it I'll happily share.


I am working on a system like this for music, (scans links on YouTube Reddit SoundCloud etc) I'd be interested in reading what you have. My email is tylergetsay@gmail.com


I'm currently editing it a bit, I'll send it in a few days (feel free to remove your address from the comment, I saved it).

I also think such a system is well equipped for matters of taste, where some peer-bubbles can even be beneficial (there is so much bad art I never want to see...).


I'm interested in reading the pages you wrote! Please loop me in ;)


Find a way to break the metadata chain connecting Tor users to onion servers on the Tor network. Most specifically the 'logged in here, connection happened on the server discovered there' edition.

Not sure ethereum is good here vs. privacy-first alternatives, but also not sure if all that is needed is Mechanical Turk, or a Stallman-esque browse-by-email. Maybe a protocol where people can get paid to run an email re-mixer as an onion service?

Build a system where users get paid to click random buttons, invoking services on onion sites. The hard part would be preventing side channel attacks... same request length, truly randomizing delay between task submission and completion (or some sort of subscription to send a block of random data regularly but sometimes it's user-controlled), no permanent logs vs. getting paid, etc.

Riffing on a few useless ideas here on a low-traffic day since the ideas are not going anywhere without execution.


I am not that familiar with Tor. Maybe someone in this thread will experiment with this idea.


Implement microtransactions. This is a good 2nd business (for mixers!), where you have existing customers who might be willing to pay a bit extra to be able to support their favorite creators. It gains extra potential if you act as an exchange, giving a currency that has a solid history of cheaper transactions.

You might gain additional leverage by being willing to collect for people who aren't connected, but that can get awkward (allow donating to charity). Also, taking care of getting funds to designated recipients who want nothing to do with ethereum basically​ means the legal setup reporting money transfers.

I think the window for this is closing as some mainstream currency is going to do this right, or some cheap-transaction currency will become mainstream/accessible enough.

I am posting ideas as separate comments. Hope the weekend crowd can give you something useful.


Summary of submitted ideas:

* Digitally verified lab book (a wiki style update/edit verified by 'signing parties') [a_bonobo]

* GoFundMe-like smart contract [j_s]

* Polling system for tamper proof results on opinion polls (for media houses) [apexkid]

* Trend Analytics of crypto-currencies on platforms such as Twitter/reddit. [AznHisoka]

* link aggregation + hierarchy of ratings from link curators to avoid suggestion engine clickbaits and sockpupetts inflence [anotheryou]

* Micro-transactions for existing customers who want to support their favorite creators. [j_s]

Good suggestion:

* Read/listen/watch to learn the basics (in a domain) in order to unleash your creativity [fazkan]

Thanks everybody for your time and consideration. At the very least, I wrote down more questions of my unknowns, what-ifs, and basics to read/experiment more with. This means a lot :)


Common man you dont want others ideas, just practice writing 10 ideas a day that can solve some problems. Dont filter while writing.

Also word to the wise, if you are asking other people for ideas to implement in a space (ethereum) so that you can learn the technology, implies that you dont know much about the field or have clear basic concepts. For now I would just read/listen/watch to learn the basics in order to unleash my creativity...

your sincerely, nobody...


Thanks for the advice. Going back to the basics is almost always a good idea before forcing down the trees. Will do.


Do what interests you. What are you interested in? Fish, sports, books, travel, food etc. Choose a domain first, then select something to apply to that.


You may have meant 'domain' as in an area of interest, in which case this is good advice. For years though, I was a proponent of the 'buy a domain name first' method of idea validation. As a result I spent far too much time and money on domain names when I should have been testing the idea/market.


Thanks for pointer. Quick summary of my historical interests: breakdancing -> Finance/investing -> automating stuff with code -> software projects -> cryptocurrencies -> functional programming. The domain I have invested in (probably will for another 3+ years or even more): cryptocurrencies; there is a lot of hype right now but I have been following the space for a while and my interest hasn't decreased and that's certainly not due to the prices. That's why I'm searching/exploring further.


Trend analytics on how often the various cryptocurrencies are mentioned in Reddit.

Sensible price alerts to prevent people from obsessing over their stock or crypto portfolip (ie an alert for BTC is in the 12k range as opposed to a daily price alert)


Make torrents editable. (eg. one torrents contains discography of a band. band releases new album. torrent owner adds one folder to the torrent which gets distributed to all the peers).

Someone will do this, mark my words.


Find a problem that does not have a solution yet. Make a solution to solve the problem and offer free for life licesenes to companies that want to beta test it and give feedback and bug reports.


Thanks for the input Orionblastar.


Find a way to overlay an ERC20 type system on top of a zero fee coin like XRB, IOTA, or other. Not what you asked for but I'd love to see it!




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

Search: