#1: You do not want randomness. You may believe you do until the Titanic crashes into your front yard and your significant vanishes into thin air. You want quite a lot of predictability, up to a degree where it might not even matter if things at the lowest level of existence are not perfectly deterministic.
#2: What's so bad about thinking about life as an exciting rollercoaster ride? The tracks are laid but the ride is still fun.
If everything is deterministic, i.e. determined, there's no free will, so you/I are just a NPC. I prefer to live in a universe where my conscious decisions matter, or at least can't be predicted beforehand.
Randomness doesn't imply free will. What if you/I are NPCs that just roll the dice before doing something? It's not you that chose the outcome, it's the dice, i.e. the laws of physics.
I don't know how free will could actually work with any kind of universe governed by a set of laws, whether they include randomness or not. So I don't believe in it, but of course in my day to day life I act as if it exists.
Yet I don't know how qualia or subjective experiences could actually work with any kind of universe governed by a set of laws, whether they include randomness or not. But I believe I have this subjective view of the world that doesn't seem to be explainable with a set of equations.
So it's weird. At least philosophy and science agree that.
Don't disagree but being the social animals we are, images and videos will never not be important. Things will always feel better when I can connect it with a friendly face.
The source, personal significance, and intent of images and videos will matter a lot, though. I'll cherish photos of my family members forever, regardless of technical excellence.
Or a photo of my freshman dorm room during exam season. Subpar image quality, lousy lighting, etc. but so many memories, positive and negative, are elicited by that fleeting glimpse from an era of excitement, boredom, stress, uncertainty, and optimism, not knowing where I was going in life, when I'd ever look back at that snapshot, but deciding on a whim to grab it during a break from cramming topics now long forgotten.
But I roll my eyes at the idea of injecting my likeness into a short clip depicting random over-the-top action sequences, no matter how photorealistic, because I've never wanted to do that.
Anybody else experiencing severe decline in Claude output quality since the introduction of "skills"?
Like Claude not being able to generate simple markdown text anymore and instead almost jumping into writing a script to produce a file of type X or Y - and then usually failing at that?
Anecdotally I'm using the superpowers[1] skills and am absolutely blown away by the quality increase. Working on a large python codebase shared by ~200 engineers for context, and have never been more stoked on claude code ouput.
This just feels like the whole complicated TODO workflows and MCP servers that were the hot thing for awhile. I really don't believe this level of abstraction and detailed workflows are where things are headed.
Claude Code became almost unusable a week ago with completely broken terminal flickering all the time and doing pointless things so you end up running out of weekly window for nothing.
I guess OpenAI got it right to go slower with a Rust CLI. It lacks a lot of features but it's solid. And it is much better at automatically figuring out what tools you have to consume less tokens (e.g. ripgrep). A much better experience overall.
I have also anecdotally noticed it starting to do things consistently that it never used to do. One thing in particular was that even while working on a project where it knows I use OpenAI/Claude/Grok interchangeably through their APIs for fallback reasons, and knew that for my particular purpose, OpenAI was the default, it started forcing Claude into EVERYTHING. That's not necessarily surprising to me, but it had honestly never been an issue when I presented code to it that was by default using GPT.
I've noticed this with Gemini recently - I have a task suited for LLMs which I want it to do "manually" (e.g., split this list of inconsistently formatted names into first/given names and last/surnames) and it tries to write a script to do it instead, which fails. If I just wanted to split on the first space I would've done it myself...
Not since skills but earlier as others have said I've noticed Claude chat seems to create tools to create the output I need instead of just doing it directly. Obviously this is a cost saving strategy, although I'm not sure how the added compute of creating an entire reusable tool for a simple one-time operation helps but hey what do I know?
it's been doing this since august for me. multiple times instead of using typical cli tools to edit a text file it's tried to write a python script that opens the file, edits it, and saves it. mind-boggling.
it used to consistently use cli tools all the time for these simple tasks.
It looks nice and I really want to engage with the page further but since my time is limited today and I'll have forgotten about this by tomorrow: What's the tl;dr?
Claude to the rescue: This is an interactive data story from The Pudding about research showing that talking to strangers makes us feel better, despite our expectations.
The piece follows conversations from a study of nearly 1,700 video calls between strangers with different backgrounds (age, race, politics, etc.). While people predicted they'd have negative experiences talking to strangers, the vast majority actually felt better by the end of their 30-minute conversations - regardless of how different they were from each other.The story argues that we've lost "bridging social capital" (connections with people unlike us) and explores how this contributes to declining social trust.
It ends with a personal reflection on helping a bleeding teenager on the subway, suggesting that despite our fears, most people will help strangers when needed - and that these connections are crucial for tackling big societal challenges.
How do you deal with specific wording that needs to be used in certain languages (often required for legal topics) or specific brand-related messages that need to be in place without any modifications? Does a developer still have the ability to manually translate certain strings?
The best solution right now is prompt engineering: turns out, AI can be tuned to provide top quality results with the correct system prompt/few shot setup, and custom prompts can be provided in the compiler config.
Longer term, I want this to never be an issue, and I feel we'll get there together with the help from the open source community!
One simple idea* would be an option to exclude certain React elements from auto translation. That would allow users to handle these specific cases “manually” while still reaping the benefits of the automated system the other 99% of the time.
we've added support for both these cases actually! :)
1. `data-lingo-skip` - excludes a jsx node from i18n
2. `data-lingo-override-<locale code>` - overrides version in <locale code> language with a custom value
3. also `data-lingo-context`
Congrats on the launch! How does your PDF to text feature compare to services like LlamaParse and 2markdown.com? Especially in terms of layout and image understanding. And for a privacy focused service, it would be great if your privacy policy would contain text :) Is the service hosted in the EU?
reply