New tiny package: github.com/AnswerDotAI/shot2… Press a hotkey, drag over any region of your screen, and whatever's written there is on your clipboard. Including text you can't select: screenshares, videos, images. Setup is `pip install shot2text` + 2 config lines. The whole library is under 50 lines of Python. That's possible because of two new @answerdotai libs, macmage and imp, which open up macOS scripting from Python. What they do, in 🧵

Very interesting paper. It is framed as: model intelligence is inversely proportional to # of human interventions However, I think that having a human might give a higher quality result, even if you could do it with less interventions. I used Solveit to implement & score myself on one of the tasks (very easy & reproducible paper setup, kudos for that! It was a relatively simple data anonymization task, my score was perfect (same as the AI). The resulting code was much simpler, almost half the LoC and methods. The median score of GPT 5.5 is 0.763. I would love to see more people's results! If you are interested in trying out a task, you can just use my dialog which is fully setup and just choose a different task! share.solveit.pub/d/62a08cb8… Share the results if you do

I'm happy to say that today I'm joining @answerdotai ! I am happy at a technical, personal and spiritual level. On a technical level, they write fantastic software (SolveIT, FastHTML, nbdev, fastspec), articles (why you shoud write) and great courses (fastai, solveit). I see no end to my learning and path to mastery in here. On a personal level, I share interests with many of them beyond sw like sci-fi, chinese or robotics. On a spiritual level, AAI's stated goal is human flourishing, but in all fairness it seems to be more or less the goal also for all companies striving for AGI. So why AAI? When I read Huxley's "End & Means" I realized that the ends cannot justify the means. Moreover, using means opposed to your goal are unlikely to bring that goal to happen. Violent revolution for example, has rarely brought subsequent peace the book argues. Many labs have great goals too like human flourishing, long-term benefit of humanity, or improve billions of lives. However, often the means seem to sue each other, refusing to hold hands, or benefit humanity (but not China). Or lock in people to their sw, sometimes leveraging dark-flow patterns or fully gatekeeping access. I assume it's very hard to run such behemots in a fully consistent way. That's not a problem for AAI. It is a small team and their founders fully embody the company ethos. @ericries taught a whole generation of founders how to build w/ the Lean Startup. However he also saw that not all means are created equal: not all ways to make money are equally good. To make sure money making would be aligned with humanity's benefit he went out of his way with projects like the long term value stock exchange. More recently, his new book Incorruptible is literally a manual to make sure that companies don't go from creating value to society to extracting value from it. @jeremyphoward constantly works to enhance human creativity. He has pushed for making AI accessible to everyone through fastai courses and making sure no ludicrous amounts of hardware are needed to follow them or do meaningful research. The SolveIT discord is also one of the nicest places on the internet. The only rule is to try to be as helpful as possible, and he is the best example of that. I once heard Alan Watts say that there's no point in planning the future, if you can't enjoy the present. It seems to me the same is true for ends and means. Too many people seem to be living in borrowed time until AGI arrives disregarding their current learning, skills and also enjoyment. If you are one of those and you are not happy with that state of affairs, you are not alone! reach out

Today I finished learning my first 1500 Chinese characters! (actually it was on November, but didn't get around writing about it until now). I took the chance to write a far too long review on my learning journey so far. I don't think anyone should read it unless you have a lot of free time or really want to learn a new language (for those who won't heed the warning, the link is below) It is titled: Excuses, Friction & FastHTML: My journey learning 1500 chinese characters because it reflects the most useful lessons learned for me. For context, I'm learning Chinese for family reasons. It's going to be a lifelong project but I only have (or want) to invest 5-30min a day tops. So it is a marathon for me, not a sprint. With this in mind, I set out to find an easy way to learn w/o too much effort. I tried a variety of approaches, like straight our learning the Dao De Jing, immersion in Taiwan, brute-force anki, and audio lessons. Success was mixed, learning happened, but not sustainably which leads me to the first word "Excuses" If you find yourself making excuses not to do it, it's probably because it's not a good method for you. I felt "lazy" because I didn't want to put in the effort, instead I procrastinated by reading how others learned Chinese. Or I kept trying new approaches w/o completing any. However, by being lazy and using that as a guiding light, I kept exploring alternatives until I found a talk by Jeremy Howard that lead me to the wonderful book "Remembering Traditional Hanzi Characters" by James. W. Heisig. ( @jeremyphoward thank you so much for that!) This book is a perfect match for my learning approach because it teaches you characters in a very approachable way, it decomposes them into smaller chunks, gives them visual stories and weaves them into a network that makes learning much easier. If I had "powered through" the other approaches I'd never have found this blessing. I had a math teacher that used to tell us that lazy ones are the clever, because they find better methods to work less by making things easier. Amen. Which leads me to the second word "friction". The method was right, now I needed to just do it every day. The only important lesson here is to minimize the friction as much as possible so it becomes a habit. For example, I got the ebook version so I could review anywhere, I got the anki app for review, I got anki cards for the book so I could just add them easily. If I could write them with pen & paper great, otherwise just use your finger. Setting a very low bar, contrary to some opinions is great. I committed to every **just** review the pending cards. If I had time and energy, I would learn some new ones, maybe 1, 2 or 5, rarely more than that. Thanks to compounding this eventually led me to cross 500, and then 1000 chars. At this point the habit was so deeply ingrained that I found myself doing up to 25 chars/day w/o any extra effort. Once I finished the book, I started trying to read simple stories, or talk to my wife. Looking up words in the book index is super annoying, but it was manageable during the learning. With the book finished, that became a source of friction which led to me... FastHTML. I ended up creating my own lookup page with all the book characters. That way I could look up by char, meaning, or number in the book. Moreover, each char had hyperlinkg to its constituents making review and lookup a pleasure, instead of having to browse through the book (best case) or navigate the ebook (even worse!) when I didn't have the physical copy. Eventually I wanted to learn words with multiple chars, so I added the feature to lookup multiple chars, and then have Gemini explain me how those chars with the meanings attributed to them by the book, gave rise to the final meaning. Absolutely perfect. Sometimes it seems a "waste of time" or "procrastination" to work on the tools or methods, but the opposite is true. The better the tools the easier the process. At the time of writing, I'm connecting the FastHTML app, so any compound word can be added directly to my anki decks, I've transcribed Pimsleur audio lessons so I can review them using the books characters or pop them into ChatGPT Voice mode to practice with the right context. Happy learning, I'll report back when I finish the next 1500.

SolveIT by @answerdotai is by far the best reading experience I've had. Their dialogue approach means you can split the text at any point and just ask questions. You can also run code if you need it, and show/hide selectively stuff to the AI (so it does not get distracted with your rabbit holes). Here's a very basic example: gist.github.com/kafkasl/d76e… People are lately reading full form books in a number of different personal approaches