Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Ask HN: What interesting problems are you working on?

news.ycombinator.com

Post reply on HN
Print view View original

51–60 of 633 posts

Re: Ask HN: What interesting problems are you working on?

#51

Post by ultra_nick »

We're trying to automate tutoring. As a college student, I spent a lot of time trying to fill knowledge gaps with textbooks, Google, or other students. Now, we're trying to provide a streamlined and customized AI Tutor service to help everyone learn faster.

It's a lot more technically challenging than we thought. For example, what's the best way to statistically guage a student's knowledge level? How do we recognize when an ngram counts as a new concept?

We haven't launched publically yet, but we have a LinkedIn page of you'd like to see when we do. https://www.linkedin.com/company/conceptionary/

Re: Ask HN: What interesting problems are you working on?

#52

Post by EForEndeavour »

travisjungroth wrote: post #10

I’m working on the expert elicitation of priors. We’re making web UIs to better capture human judgement, especially for forecasting. I work at a data driven company, sometimes I think a bit too data driven. I want to capture more intuition for our systems in a rigorous way. It’s incredibly interesting and very niche. There are a handful of academic papers out there. I think it will be a big deal in the next 10 years.

This sounds incredibly interesting to me!

For years I've had this idea rattling around in my head to build a web UI that shows a graph of some historical quantity, e.g., company revenue, disease cases, global mean temperature anomaly, stock prices, etc. The historical data is shown as a curve which terminates at the present, and there's some empty space for the future. The user then uses a freehand drawing tool to extend the historical curve into the future, based on what they feel will happen. Save this data, and let users view some average(s) of all crowdsourced curves, grouped by submission date. Stretch goal: let users track their own forecast accuracy, and let users view forecasts weighted or filtered by the historical forecast accuracy of contributing users.

Is this close to what you're working on? I have no idea how useful or interesting this would be, but I feel like I'd learn a ton just trying to build it. Alas, I just haven't made the time, and lack the skills to hack something together in a reasonable time.

Re: Ask HN: What interesting problems are you working on?

#53

Post by petercooper »

throwawayacc2 wrote: post #2

One of my hobbies is tax optimisation and finding and exploiting tax loopholes. I’m a contractor, I work through my company. It’s way more tax efficient than being perm. But, it could be even more efficient. So, I did some research. Turn out, where I am, a business can have multiple business codes. Activities the company is registered as performing. Now, I was and still am a software development company. But I also a…

"A person after my own heart," I originally thought. I, too, enjoy reading HMRC technical guidance, although I haven't actually taken it to such extremes, instead focusing on genuine deductions few people tell you about (like annual medicals).

Your post was entertaining, but there might be a fly in the ointment! Are you actually carrying on a trade? Your consulting work is trade, but there are rules around businesses having multiple lines of business and what is and what is not considered a trade (see BIM20090 and possibly BIM85740). If your art is not commercially available and making more money than it "costs" in expenses, HMRC would probably determine that this line of business is a hobby with no expenses deductible.

If, however, you are selling this art, it's a fantastic wheeze, and also a quite legitimate one as you would, indeed, be a professional artist. You could, too, perhaps find ways to use said art in your trade such that it would commercially justify its creation, even at a high cost.

Re: Ask HN: What interesting problems are you working on?

#54

Post by zciwor »

My sister works as a veterinarian and is experiencing some serious burnout from her current practice (a subsidiary of a larger corporation).

I guess it's pretty common practice for corporations or private equity to snatch up smaller independent practices and then run them like a sweat shop. Her current CEO's resume is a list of Starbucks, Walmart, and a few other corporate roles that are totally unrelated to the medical industry.

We're tinkering with a platform that can somehow disrupt this cycle and add some transparency to the practice structure for the would-be employee.

The current form is kind of a niche Glassdoor, or we've been thinking of it as the "online dating between veterinarians & practices." Still playing with the concept.

Anyone have any thoughts on how to take out these massive PE firms that are plaguing the Vet Med industry?

Here's some more info on the problem: https://www.ftc.gov/news-events/news/press-releases/2022/06/...

Re: Ask HN: What interesting problems are you working on?

#55

Post by snicky »

throwawayacc2 wrote: post #2

One of my hobbies is tax optimisation and finding and exploiting tax loopholes. I’m a contractor, I work through my company. It’s way more tax efficient than being perm. But, it could be even more efficient. So, I did some research. Turn out, where I am, a business can have multiple business codes. Activities the company is registered as performing. Now, I was and still am a software development company. But I also a…

It's all fun and games, but can't the auditors just challenge your "artistic production" by asking for the actual invoices proving any sort of income from it? I mean lots of costs and no related income is questionable AF. Or do you actually sell any of this "art"?

Re: Ask HN: What interesting problems are you working on?

#56

Post by ultra_nick »

zciwor wrote: post #54

My sister works as a veterinarian and is experiencing some serious burnout from her current practice (a subsidiary of a larger corporation). I guess it's pretty common practice for corporations or private equity to snatch up smaller independent practices and then run them like a sweat shop. Her current CEO's resume is a list of Starbucks, Walmart, and a few other corporate roles that are totally unrelated to the medi…

My wife and I are suffering personally from this. I'll help for free.

I'm convinced Vetmed will follow human medicine's history. Currently, the future of human medicine is small clinic subscription medicine without insurance. Apps like Roo could help with temporary staffing issues.

Let's talk: https://www.linkedin.com/in/nicholascgilpin

Re: Ask HN: What interesting problems are you working on?

#58

Post by qsort »

omeysalvi wrote: post #33
qsort wrote: post #17

I'm on a (currently very experimental) project where we're trying to answer natural-language queries using data from an arbitrary database (given hypotheses on the schema). The natural-language plaintext is fed into a language model that annotates it. That output is converted into an intermediate representation based on relational algebra, which we manipulate and finally compile to SQL. It's probably not scientifical…

Are you working on the Yale Semantic Parsing and Text-to-SQL Challenge? Are you already on the leaderboard? The topic has always fascinated me and got me learning natural language processing to understand how it works.

We are going for a narrower scope than that. We are working with stricter restrictions on the database schema, where you are only allowed to join on synthetic keys and there is an enforced distinction between measures and dimension attributes.

On the other hand we are aiming for reasonable response times on large scale datasets.

Re: Ask HN: What interesting problems are you working on?

#59

Post by olinda »

zschuessler wrote: post #5

I heard few song lines from a new artist I've discovered. It has me thinking on a problem I want solved. Although, I'm not actively working on it. Why aren't there more employee-owned companies? And I don't mean solely shareholder programs, I mean actually owned by every employee and they are paid dividends of profit after all planned R&D costs. Similar to the Alaska oil pipeline bonus. Even interns. All studies show…

How would this be different from a partnership like a law firm or consultancy? Or are you just suggesting that more firms are structured as partnerships?

Re: Ask HN: What interesting problems are you working on?

#60

Post by anthonyhn »

I'm trying to improve my search results by building my own Search Engine.

Over the past couple of years, the popular search engines haven't given me great results, and I've had to use a series of different engines, search operators, and dorks to prod out the search results I'm looking for, with mixed success. The front ends have also become a bit too feature rich for my use cases; I prefer a minimal front end with just search results, but most front ends have links to news, videos, drop downs, cards, cookie banners, SEO spam, and have too much javascript. Too much time spent loading and rendering the page, and too many requests, in my opinion. Instead, my search engine only take 1 HTTP request per search result page and is only a few KB in size.

Right now this project is still in the early stages, and mostly proxies bing's api with some minor adjustments, but if anyone is interested in testing it out here is the link to the search engine [0] and a link to the blog with an RSS feed to follow future updates [1].

[0] https://simplesearch.org

[1] https://simplesearch.org/blog.html

Post reply on HN