I built a chatbot to fight loneliness and social isolation for seniors. It started with my grandma. She doesn't have a smartphone and internet, so it basically transforms photos into real postcards. She receives it directly in her mailbox, and it makes her really happy. The chatbot also reminds me to send when I did not, so she keeps updated regularly.
I released it to the public last week. you can find it here https://postcard.im (the link open a fb messenger conversation with the chatbot)
Ask HN: What interesting problems are you working on?
561–570 of 709 posts
Re: Ask HN: What interesting problems are you working on?
#562I am working on a natural language parser using symbolic AI (no machine learning...). It's working a bit like a multi-pass parser for programming languages, but with the ability to handle multiple ambiguous meanings of a sentence at the same time. An English sentence is translated into an intermediate representation. Or rather, depending on the complexity, hundreds or thousands of intermediate representations for the…
I’ve been working on merging statistical deep learning with a knowledge graph like representation of the data. I run a consumer site focused on privacy and build nlp secret sauce underneath to serve users. Ping me at hn username @doerhub.com if you want to chat.
Re: Ask HN: What interesting problems are you working on?
#563Creating the worlds best IP address and domain name APIs and data sets, at https://ipinfo.io and https://host.io . We've solved scaling and reliability (we handle 20 billion API requests a month), and we're now focusing almost all our efforts on our data quality, and new data products (like VPN detection). We're bootstrapped, profitable, and we've got some big customers (like apple, and t-mobile), and despite being a…
Why/how is this better than existing IP solutions (e.g. https://www.maxmind.com/en/home or https://www.digitalelement.com/)?
Re: Ask HN: What interesting problems are you working on?
#564Earlier quoted context omitted.
Maybe have a dispatch interval. The weekly-on-tuesday emails get sent one every 5 minutes starting at some time.
Very good. I’m noting all these suggestions down. The app only launched a few days ago. I wanted to make sure it was a valid product first before doing too much to it. I’ll gradually add more functionality and examples over time.
Sounds like you have your mvp and have an incremental plan going forward. Good. The thing is out in the world - that is something a lot of folks don't do.
Re: Ask HN: What interesting problems are you working on?
#565I'm growing the freshest lettuce, iron-rich kale, and a lot of other leafy greens! While in college (CS & Math), I got heavily interested in growing food in the most efficient and healthiest way possible. I was a dreamer when I started so I thought more of how to grow 'earthly' produce on Mars, but then I realized that my own planet Earth is so massively underserved. It's basically like this- I mastered growing leafy…
Amazing! It would be awesome if people living far from traditional agricultural areas could access fresh greens without insane transportation costs (both financial and environmental). Are your farming systems fully automated? If so, has that been more of a software challenge, or more of a mechatronics challenge?
>It would be awesome if people living far from traditional agricultural areas could access fresh greens without insane transportation costs (both financial and environmental)
That's what actually got me started. A head of lettuce on average 1200 miles (https://ucanr.edu/datastoreFiles/608-319.pdf) and it is so disconnected from the site of consumption.
My vision is to have distributed farms (as opposed to conventional wisdom, i have found that smaller indoor farms will be more profitable) every eight blocks or so.
Not really- It's quite manual (as of now). I had to change my country almost three times since I started; so I'm rather focussing more on data, and training algorithms part to figure out the right parameters (and the farm is a just a testbed). One example would be to have a $5 camera for measuring growth than buying a $100 3D what not camera.
Re: Ask HN: What interesting problems are you working on?
#566I'm growing the freshest lettuce, iron-rich kale, and a lot of other leafy greens! While in college (CS & Math), I got heavily interested in growing food in the most efficient and healthiest way possible. I was a dreamer when I started so I thought more of how to grow 'earthly' produce on Mars, but then I realized that my own planet Earth is so massively underserved. It's basically like this- I mastered growing leafy…
> THE FINAL PRODUCE IS THE ULTIMATE MANIFESTATION OF THEIR PLATONIC IDEAL FORMS How's the taste? Not denying it's possible to grow food very efficiently indoors but a vastly oversimplified opinion is that plants need sunlight to be tasty. Is this wrong?
You'll have to buy my words- but taste wise (based on my surveys too) it's the 'best' they have had (mostly city dwellers I'm talking about).
Yes, you don't really need sunlight whatsoever. I was myself shocked until I recalled high school biology concept of genotype and phenotype i.e. the genetic structure that manifests itself given the right physical conditions (at least of plants.) As for the plants' nutrients, here's a classic- Teaming with Nutrients: The Organic Gardener's Guide to Optimizing Plant Nutrition, by Lowenfels. I was amazed to find how complex, yet simple plants are.
Re: Ask HN: What interesting problems are you working on?
#567I'm growing the freshest lettuce, iron-rich kale, and a lot of other leafy greens! While in college (CS & Math), I got heavily interested in growing food in the most efficient and healthiest way possible. I was a dreamer when I started so I thought more of how to grow 'earthly' produce on Mars, but then I realized that my own planet Earth is so massively underserved. It's basically like this- I mastered growing leafy…
> mastered growing leafy greens in indoor closed environment I love this! Makes me happy to see someone's working on such an interesting problem that would benefit many. For feedback, I believe using photographs of the leafy greens would be effective in communicating your vision.
Thanks I'll do that.
I actually graduated from college this year; and for personal reasons I've had to change countries; now I'm in another Master's program... ready to drop out anytime. The whole project has been dead for months once in a while! I'm more trying to leverage ML for optimizing things. I guess that's what modern farming is missing (not ML per se, but optimization).
I'm trying to raise some investment (or in the worst case bootstrap and risk everything in the next few months), then I will go crazy with the idea.
Re: Ask HN: What interesting problems are you working on?
#568I'm growing the freshest lettuce, iron-rich kale, and a lot of other leafy greens! While in college (CS & Math), I got heavily interested in growing food in the most efficient and healthiest way possible. I was a dreamer when I started so I thought more of how to grow 'earthly' produce on Mars, but then I realized that my own planet Earth is so massively underserved. It's basically like this- I mastered growing leafy…
You should check out "The Real Martian" https://www.youtube.com/channel/UCd8t8Dq8oZeAjGDx_87azBw/abo... and Beanstalk (a YC company) https://www.beanstalk.farm/
Funny story: I was rejected by YC last batch. But I get it- I thought they look for traction and what not, so I rather made the pitch video on a very specific aspect of Hexafarms- which is monitoring, since some people were willing to check it out. No doubt YC would reject it. On the other hand, Thiel Foundation reached out to me, but they had some drop out thing and what not which I was not able to fullfil (and after a while they also stopped reaching out too).
Thanks for the references.
Re: Ask HN: What interesting problems are you working on?
#569I'm helping to build a scalable system for delivering high value and life saving medical supplies to hard to reach places via autonomous aircraft. The system is currently operating in Rwanda and Ghana, and aggressively expanding over the next couple years. Specifically, I spend a lot of time thinking about and writing embedded software. The aircraft is fully autonomous and needs to be able to fly home safely even aft…
By "hard to reach places", are you referring to difficulty due to geographic positioning, or geopolitical concerns?
Re: Ask HN: What interesting problems are you working on?
#570The environment and how to lead people to enjoy acting sustainably long-term, so they spread that joy and sustainability to others, not coerce. I've found a strategy I think believe will work -- my Leadership and the Environment podcast. Here's the podcast: http://joshuaspodek.com/podcast Here's an episode clarifying my strategy: https://shows.acast.com/leadership-and-the-environment/episo... Here's my corporate stra…
> "enjoy acting sustainably long-term, so they spread that joy and sustainability to others"
This sounds great, can't wait to check it out. After taking a Permaculture Design Course, I fell in love with the ideal of living a better, more "luxurious" life that has not only a lower impact on the planet, but a positive impact. I'm on my way down that path, but still figuring out how to make it replicable.