gault8121

The loading bar is extremely subtle and hard to see. I missed the tooltip message in the corner. It would be helpful if there was more clear visual feedback that the query is being processed or if the tooltip stayed open until the query was finished.

That's basically what I am doing with Quill.org (I'm the founder / executive director). Quill is a platform for learning tools and games - we support six web applications that plug into a common, open source platform where teachers can assign activities to students and monitor results. At the moment, our applications are all focused on literacy, and they are more "tool-like" than "game-like". However, we use AI to assess writing, and the AI creates a game-like experience of writing different responses and getting different pieces of feedback (somewhat like a MUD RPG). We're creating the applications in-house now, and the API is not public, but the hope is to open it up in the future to allow others to launch tools and game on our platform. Quill is now serving more than six million students across the United States, about 12% of all K-12 students.

Tech non-profits: https://jobs.ffwd.org/jobs

What's the rationale for not including illustrations?

These are tools for teachers to catch plagiarism. He is saying he wants better tools to detect plagiarism because these don’t handle similarity matches well.

How did the pricing compare? What was the Render server needed to match what you were using with the Heroku Dyno? I run a high-traffic nonprofit edtech app that runs on 3-5 Heroku L Dynos, and I'm curious how well Render will perform at this level.

Has anyone done a load test comparison for Heroku vs. Render.com? The "Pro Ultra" on Render is $450/month for 32 GB RAM + 8 CPU. The Heroku Performance L Dyno is $500 a month for 14 GB RAM. The Render server seems like a much better offering.

Hi there exdsq, I run a nonprofit, Quill.org, that serves five million students in the United States, and I am building an educational tool that is very similar to the "cognitive aid" goal listed as a project idea. I'm hoping to get some feedback from someone who is familiar with this space, and I was wondering if you had a moment to scan over our application before we send it? If you're interested in taking a look, it'd be a huge help if you could reach out to me at peter (at) quill (dot) org! I'd be really grateful for your input. -- The learning tool we are building now maps to this project idea from the FTX Fund: "AI-based cognitive aids - One of the great hopes for advanced AI systems is that they might enhance human reason—allowing people to explore lines of argument more carefully and efficiently, and to detect important errors in their reasoning. We’d like to kickstart this work, so that it keeps up with AI progress as much as possible. For example, could a fine-tuned version of GPT-3 be trained to identify misleading statements, and provide the best arguments and counterarguments for different views? We’d love to see products with this sort of technology that people will actually want to use."

Awesome, thanks for following up! We also use coreference resolution - we specifically found Allen NLP's coreference to be the most accurate in an analysis of a few different offerings: http://docs.allennlp.org/v0.9.0/api/allennlp.models.corefere...

This is an interesting tool! It would be cool if you could take say 20 articles from Gothamist, Streetsblog, NYT, NY Post, Pix11, etc. and see how they all ranked in this system. That could be a really interesting blog post. -- Which tools are you using for these NLP judgments? Are you using SpaCy at all? Are there any models you've built here, or is this all rules on top of an NLP model? I'm working on NLP models for education, and we use SpaCy a ton. I'm peter@quill.org if you want to learn more about how we're using this.

wow, this will go down as an all-time tech fail.

We use Moqups (moqups.com) extensively for quick scratch-pad designs. Moqups includes commenting tools and collaboration. What are the benefits of NextUX over Moqups?

The screenshot feature looks helpful, but taking screenshots isn't really a pain point for us as there are a ton of tools that do this for us and make it easy to drop into Moqups (including a Moqups chrome extension that does exactly this). If you say scrapped my entire site, took screenshots of everything, and gave me a digital copy to work off of, that would be interesting. Alternatively, if your design tools were directly embedded into the site, that would also be really interesting.

These are both big asks, and they might not be feasible, but the more you can close the gap between taking an existing site and turning it into a design space, the more attractive a tool like this would be for our org.

It would be helpful if the site could define what they mean by the average. The average visitor to the website thanaverage.xyz? The average American? the average person around the world? It seems like it would be more interesting to make a guess about the average visitor to the website, but it's not clear if that's the case or not. Compared to the average American or the world, the numbers for this small segment of people who would visit this website would not reflect the average.

Yes, that is great to hear! Pre-pandemic, Quill had the most luck with in-person hackathons, where folks could spend a day contributing to Quill, and some of those folks were then interested in staying connected with the work. Hopefully as those events return, you may be able to scale up your community building efforts.

That sounds like a great and ambitious roadmap Aly! :)

Best of luck scaling to support 100x users! In case you are not already using it, New Relic's new program for nonprofits provides their entire platform for free, and it can be really helpful for scaling by seeing how the app is performing. In case you are not already using it, here's the link: https://newrelic.com/blog/nerd-life/introducing-observabilit...

Congratulations Aly! It's Peter Gault from Quill. What's next on your roadmap? What are you hoping to achieve over the next year?

In the report, not the summary, Firebase is listed as the most used 3rd party, with Firebase Messaging and Firebase Analytics being designated as "High Risk." When the report says "these apps are sending data to Google" it sounds like they are sending ad tracking data to commercialize off of this data, but tools like Firebase are common web hosting platforms, not data tracking tools. This is like saying data is being sent to Amazon Web Services.

Here is the report: https://me2ba.org/school-mobile-apps-student-data-sharing-be...

As an edtech founder, I 100% agree that there should be more time spent holding school apps accountable for the data that they sent out, but this report doesn't accurately capture the risks present and how to hold organizations accountable.

I'm building Quill.org to help students become stronger writers and critical thinkers. We're turning the powerful writing strategies detailed in this piece into free and open-source learning tools that can help millions of students: https://www.theatlantic.com/magazine/archive/2012/10/the-wri...

We're now building new critical thinking tools that help students build logical claims, use precise evidence, and articulate their own ideas. If you'd like to learn more, my email is in my HN profile.

This implies that YC will continue to offer a remote option: "In a post-COVID world many of the changes we were forced to implement will live on just as several pre-pandemic experiences will return."

Is that what this is saying? Or is YC saying that they are going back to on-site after the summer batch?

So after reading this article, I wanted to set up some type of automated alert that would email once the rate reaches a certain threshold.

If you want to know when treasury bills reach 2%, I figured out a way to do this. My brokerage account doesn't have an easy way to set alerts on t-Bill price changes, and an extensive search on Google has not resulted in any tool that allows me to track the specific ticker (TMUBMUSD01Y) easily online.

The only thing that seemed to work was to use this website scrapping tool (https://www.wachete.com/wachets) to listen to this website (https://fred.stlouisfed.org/series/DTB1YR).

You can program the scrapper to look at the value change of the yield and set it to trigger an alert once it passes a threshold. I set up 0.5% and 1%.

I don't read financial news regularly, and it'd be helpful to get a heads up when this is happening. If anyone has a better way of doing this, please let me know.

Wow, that's really interesting. Did UpToDate publish their entire content library as of April 2018 under an open access license?

There seems to be a ton of good data here that you could use to build some type of WebMD competitor.

Yes, that's an interesting point. It seems like an individual medical provider is incentivized for repeated care where a large facility is incentivized for a fast outcome.

Yes, exactly - it's though big enterprise contracts that generate revenue, but they are hard to close as a startup. Selling this service to companies that already have those contracts is a faster way of creating value for those customers.

This article's thesis seems to be that medical professionals are not incentized to provide the best interventions, and as a result, wouldn't pay for this service. However, what the author fails to mention is the competitors in this space that are successful, such as Up To Date, which provides really high-quality research trial data: https://www.uptodate.com/home

Rather than building a product that informs medical professionals about effective interventions, I wonder if the creator would have had more success if he deeply explored what sources of information these medical professionals pay for now - do they pay for anything at all, such as UpToDate, and don't want to pay this because it's an additional expense? If the creator found which sources people are using, the creator could sell this database as a feature for these partners and widely disseminate this data through partner channels rather than creating a competing source of information. It seems to be a case of this being a good instance of a B2B2C model, where selling this service to other businesses that sell directly to medical professionals could be more viable than trying to sell directly to them.

Alternatively, if the creator wanted to sell to patients, rather than medical professionals, the blueprint here is all of the consumer reports companies, such as Wirecutter, which is one of the New York Times's most popular services. Here, again, a "Wirecutter for medical interventions" could be quite successful, and you could sell this service to media companies that provide consumer reports as a service that would bolster these companies.

It's bad the creator wasn't able to find traction, as getting more medical data into the hands of consumers could have a huge postive impact over time.

Yes, as a philosophy major, I also agree that anything that is not 20th century philosophy is not really that helpful. It can be interesting academically, to trace the evolution of philosophy, but it won't really scratch that itch of engaging in the big questions.

The biggest value I got out of being a philosophy major was building reading concentration skills - being able to read dense passages takes a lot of commitment and practice, and that can build thinking skills. The internet teaches us to skim and do fast/shallow reading, which doesn't lead to deep thinking. Reading older philosophy is a great way of building those deep reading skills, even if the actual content of those texts is not applicable to modern life.

To the OP, if you see this, I would suggest looking into Experimental Philosophy. It's a branch of philosophy that focuses on using neuroscience to understand how we think about the world. Books like Thinking Fast and Slow may be much more interesting to you.

Couldn't you "pen test" your own hiring process? Have some folks stage applications and see the process failures first hand. You might get better data from running these experiments yourself rather than through a survey.

Well it's free and in the browser? For an entire generation that never played Pharoh, this would be a lot of fun. It would be cool if the author open sourced this and people could submit their own mechanics.

This is really fun! It reminds me of playing Pharoh https://en.wikipedia.org/wiki/Pharaoh_(video_game)

A few ideas: (1) once you build everything there isn't really an end game. One easy way of adding more of an end game now is to do something like factorio and make upgrading the keep really expensive - you could have a couple of more advanced upgrades that require a lot of gold and glass for example. This would make the player need to harvest the full map. (2) deleting roads got really slow on a full map. (3) I clicked on the menu screen by accident frequently, but clicking on the menu again doesn't close it (you have to go to "return to game"). If you could make clicking on it again close that would be great. (4) the one thing that was a bit challenging is not knowing what my resource limits are. Adding resource limits as small text under the amount I have now would be really helpful.

As a future game design, it could be fun to add a public health element - homes can get the plague, and you have to build wells for clean water, morgues, parks, etc. If you build too dense, the plague spreads faster, but you also have a productive city. It would add a dynamic of having to carefully plan where you build roads and, as a result how dense you want your housing to be. A worksite could be plagued just like a house, and then wouldn't produce resources. This might work well with the current gameplay.

The video game industry takes advantage of young people who have dreams of making video games, and think that suffering through terribly paying jobs is the way to make that happen. That illusion gives cover to the industry to perpetuate terrible working conditions by abusing low paying employees and then having new young people replace them. There isn't accountability when companies do not pay a penalty for the treatment of their staff.

I am a huge Notion supporter - it's our organizaation's brain and one of our most important tools.

However, having used the tool for more than a year at a 20 person org, one of the biggest pain points I have seen is people not knowing where to place new pages. There is "Notion anxiety" about not wanting to mess up the board.

One thing we've done to address this is by adding two common tables, the Documentation table and the Meetings table. Anytime anyone wants to create a new doc, they can simply add it to the documentation table and then tag it to their team, and we have a full list of all of the docs everyone has written. From there, if a certain piece of documentation needs to be somewhat else, it can be pulled out and moved there. Additionally, each team member has a default "project board table" and if you say need to create a quick to do list, you should just add it to your personal board. Having these "default tables" rather than a special table for each specific project may really help your team.

Notion requires intentional planning, and you need to be opinionated about the architecture. If you just keep adding things on haphazardly, it's going to turn into a mess.

Finally, I do agree with OP that notifications could signicantly be improved. If anyone from Notion is reading this, please think of notifications like emails. At the moment, notifications are only in the state of read/unread, and as soon as I open my notifications, they clear. Instead, I want my notifications to be a mini-inbox, where I can view notifications, respond to notifications, archive completed notifications, and filter notifications into channels (the same way I can filter emails). Notifications are a to do list, and should be thought of as such. Having this set up would address OPs concern about being able to filter notifications at the email level by setting the filtering logic in the app.