HN user
If you want a threadwise.app premium account for free, please shoot me an email at admin at the domain name above I will upgrade you for free in return for some feedback.
You need to have one to understand the unconditional loyalty, love, and the bond. You come to appreciate it even more once you lose them. Having a dog is the best and worst thing in the world. Just lost our fur baby a week ago...
I don't like when my food is too mushy. This is a perfect ratio for me
I've been doing something similar for breakfast, one cup of oatmeal + one cup of water and about two tablespoons of chia seeds, microwave for 2 minutes. Add a banana and some honey, top it with whole roasted almonds and some raspberries. It has been doing wonders for my digestion. I'll have to try to add olive oil as well. My LDL was 150 last time I checked. I wonder what it is now since I've been doing this meal several times a week.
I am building a service to make accessing AI as easy as sending email called ThreadWise
You can email prompts directly to your ThreadWise address and get instant AI-powered responses, essentially an always available co-worker. Another great feature is the ability to schedule recurring tasks and since the AI has web access, you can get things like:
Daily mortgage rates or airfare price monitoring
Weather and news summaries
Sport scores, jokes, quote of the day
Pull data from public APIs (and more)
So you can essentially use it as a personal newsletter, crafted to your taste.
The free tier will let you test this out for free! I am looking for some feedback/criticism, testing, and additional ideas and I am open for collaboration if you have experience with sales. Also open to hearing which scheduled tasks people would find most useful.
Why I built it: I noticed a trend online, as well as with family/friends, that people would like to have a quick access to AI in instances where they couldn't always install apps or use browser-based tools (such as in remote/low bandwidth environments). This is when it him me, email clients already have all the features needed to interact with an AI (text + attachments) and I quickly got to work.
Some of the advantages are also that since there are no new apps, or browser tabs needed, the tool is ideal for companies who don't have the bandwidth to setup full fledged AI solutions on their own. The companies can choose either between public LLMs (e.g. OpenAI) or host everything on-premise with locally run models, so no data ever leaves the premises.
Eager to hear what you all think!
I'm building something similar. See my comment the OP above:
I'm building something similar. See my comment the OP above:
I'm building something similar and related to the other comments below! It's not production ready but it will hopefully be in a couple of weeks. You guys can sign up for free and I will upgrade you to the premium tier manually (premium cannot be bought yet anyway) in exchange for some feedback:
Yeah, OpenInterpreter does this (you are not limited to OpenAI only): https://github.com/OpenInterpreter/open-interpreter
I wrote a wrapper around it that works in a web browser (you'll need an OpenAI API key): https://github.com/uhsealevelcenter/IDEA
Thanks for sharing this! That was the exact info I was looking for, didn't know the wholesale price was so low. But it does make sense that transmission and storage is what is inflating the retail price.
This is a cool concept and I love the idea but the math on the money earned from the 3276 solar panel doesn't add up. The article says the farm owner makes about $20,000 a year from the solar farm.
I'm assuming that each solar panel is 2 by 1 meter, which would mean that it produces about 400 watts (20% efficiency at 1000 watts per sq meter coming from the sun). You can use this calculator to estimate how power you can produce at the given location for a given system size in kilowatts: https://pvwatts.nrel.gov/pvwatts.php
The system above is 1310400 watts or ~1,310 kW, which according to the calculator produces about 2 million kWh/year.
If he makes $20,000 that would mean that he gets paid only $0.01 per kW of power. And even if my assumption above about the size of each individual panel is off by a factor of 2 and they are only 1 sq meter in size (which I think they are not because the article states that the solar farm can power about 300 average households, which require the annual power output to be more than what I estimated above) that would make $0.02 per kW of power. How is it possible that the amount earned per kW is so low when the utility companies in Colorado charge about $0.14 per kW (effective rate)? And who is actually the customer here and where is the money coming from? I'm just curious to learn more.
Also the AI generated images that can't get the fingers right. Have you ever tried to look at your hands while lucid dreaming and try counting fingers? There are some really interesting parallels between the dreams and diffusion models.
Here is a decent explanation as to why not: https://youtu.be/CHJKKsOHtAk?si=FYd1FawD-_GN_Cex
There's a video poker version, which I played on a machine years ago, where after being dealt a winning hand we have an option to infinitely double the winnings if we correctly guess if the next card is higher or lower than 7. Does anyone know what I am talking about and what the name of this poker variation is called?
Or could just use a phone/tablet as a "viewport". I know it wouldn't be as immersive but the barrier to have it adopted would be a lot lower.
Not even close to what we as humans emit. The rate of the global annual natural CO2 emissions are 1-2 orders of magnitude slower than anthropogenic emissions.
The photo doesn't do it justice tho. Just like a photo of the full Moon low on the horizon doesn't look anywhere close to what my eyes are seeing. Now imagine a full moon, with a diameter roughly 4 times as big, that's how big the Earth is looking back from the moon. North America itself is the size of the full moon. Pretty wild to think about.
"The Moon diameter (approximately equal to the distance from New York to Las Vegas)"
Imagine how glorious and huge the Earth must look like from the Moon.
My wife and I made a small, word based multiplayer mobile game (gibberish game) in 2020 during first lockdowns. It's been making over $500 consistently ever since: https://apps.apple.com/us/app/guess-the-gibberish/id15137961...
Were the interviews not technical in nature?
You are right, Earth does have mechanisms through which it regulates the CO2 and tries to bring it back to "balance". But it is important to note that these mechanics (e.g. increased in weathering) take 10s-100s of thousands of years to take effect.
I was a lead developer for this fun project. The USGS flew a drone with a LIDAR over a national park in Hawaii to create a Digital Elevation Model (DEM) of the park and I used different GIS tools to convert and show the DEM as a 3D model in Unity3D. The sea level rise scenarios are based on NOAAs scenarios and local factors (more details in the app).
Conveying the urgency and the real risk posed to coastal communities due to sea level rise is immensely important, not only to climate researchers but to laypersons and general population directly affected by the effects of climate change. Standard scientific materials and tools for depicting sea level rise projections, such as time series or static maps, often fail to communicate the immediacy and the real threat posed to local communities and infrastructure. So the main goal of the app is to help general public conceptualize the threat imposed by sea level rise.
In future, the method could be scaled to be used at any coastal location (as long as we have the necessary data). If there is interest, I will go into more details on the data workflow of how the app was made. Android version should be live soon (under review).
One of the feature that this is missing is a plain 3D viewer (without the need for camera and augmented reality). And maybe a screenshot option to take a photo of the terrain and the impact. Probably going to be available in the next version.
Feel free to ask any questions.
Since the pandemic started my wife and I had some extra time on our hands and have decided to build a couple of mobile video games. The one posted here is a result of us going to a museum and playing a physical board game that we couldn’t solve after sitting in one spot for close to two hours (almost didn’t check the rest of the museum out).
I searched the App Store for the puzzle but couldn’t find It and that’s when I decided to make a prototype on my phone. Once I built it, it took me a couple of days to solve it. I then showed it to a couple of my friends and they were immediately intrigued as the puzzle looks easily solvable, but they would spend an hour or two playing it on my phone (even though I begged them to return it back to me). This prompted me to go ahead, polish it, add new levels, and make it publicly available.
I need some advice on how I could automatically find am optimal solution for the puzzle. Is the A* my only option? The game was built in Unity. Does anyone know of a way to use Unity ML agents to find a solution to the puzzle? I would like to be able to show users hints as well as maybe even automatically create puzzles (would check if a solution exists and how difficult the solution is). Any other feedback more than welcome.
The easiest way to learn unity (or any other new framework/engine/language) in my opinion is to recreate the project that you've already built. And you have one.
The curve is a bit steep but once the things start clicking you productivity goes through the roof. You'll be able to make game prototypes in a few short hours.
I admit that building 2D games with unity is a bit of an overkill but the fact that everything just works across different platforms makes it well worth it. I originally developed games using Java and a cross platform framework called libgdx. It was fun till you needed to implement things like cross-platform advertisements and IAPs. With unity you know that the support will be there for the foreseeable future and no other engine comes close in terms of the available free online resources.
Congratulations! Game development is what got me interested in programming in the first place. Once I put my first pixel on the screen and I figured out how to interact with it blew my mind, limitless possibilities. Since the pandemic started my wife and I have published one game [0] and currently have a second one [1] in a pre-release phase (probably going to make a separate SHOW HN post once it is live).
Keep up the great work! We use Unity3D for our development
[0] https://apps.apple.com/us/app/gibberish-game-against-friends...
[1] https://apps.apple.com/us/app/the-setting-sun-block-puzzle/i...
Building a puzzle board video game (inspired by a physical wooden board puzzle) using bits to represent the board (Bitboard game design [1]). Here is a screenshot for one of the levels: https://imgur.com/a/cLIsYII
The goal is to slide the red tile where the yellow ones are. I am trying to develop an algorithm that would be able solve these type of puzzles automatically so I can provide user with hints. Can anyone provide any ideas or resources on how this could be achieved? If you look at the screenshot you'll notice that the board always has two empty cells that you can slide the tiles into. You can also pre-order the game here: https://apps.apple.com/us/app/the-setting-sun-zen-puzzle/id1... I am planning on releasing it by Thanksgiving.
[1] http://eprints.qut.edu.au/85005/1/__staffhome.qut.edu.au_sta...
Is there a reason your minimum targeted iOS version is iOS 15? Are you using any APIs/features that are only available in the latest iOS?
Kind of related. I used similar data to produce an Augmented Reality version of the (almost) entire Hawaiian Island chain with both topography and bathymetry:
I think you are confusing gigatons of carbon (C) with gigatons of carbon dioxide (CO2).
For a second I was wondering why don't they just apply the same technique on humans and train their model that way instead... that's when I remembered that it involves sticking electrodes into one's brain to get the necessary data. What are the chances that we'll be able to read the brain activity noninvasively (thorough the scull) and at detail great enough to achieve similar things?