afro88

31 ms·

301.
by afro88 2y ago
The dystopian bit is the corporation having this data, not you. And having this data for all users. Pretty much everyone has a mobile phone.
302.
by afro88 2y ago
For those that prefer text: it seems they use a weaker but faster model for the "predicted output" / speculation. Pretty smart.
303.
by afro88 2y ago
You can imagine a conversation with an LLM getting to that territory pretty quickly if you pretend to be an unfair tax collector. It sounds impressive on the surface, but in the end it's all LLMs talking to each other, and they'll
304.
by afro88 2y ago
Perhaps I've made a big assumption / oversimplification about how this works. But.. > LLMs are stateless and they do not remember the past (as in they don't have a database), making the training data a non-issue here Yes.
305.
by afro88 2y ago
> this paper seems indicate that election-related activities were coordinated by groups of AI agents in a simulation I mean, that's surely within the training data of LLMs? The effectiveness etc of the election activities is likely
306.
by afro88 2y ago
You just fix the mistakes and keep moving. It's like autocomplete where you still need to fill in the blanks or select a different completion.
307.
by afro88 2y ago
So more or less on par with continue.dev using a local starcoder2:3b model
308.
by afro88 2y ago
For anything complex, move in small steps. For anything truly novel, or on a codebase with a very bespoke in house architecture or DSL, yeah you won't get much out of it.
309.
by afro88 2y ago
> But the idea of letting an LLM write/move large swaths of code seems so incredibly irresponsible. I think this is where the bimodality comes from. When someone says "I used AI to refactor 3000 loc" some take it to mean t
310.
by afro88 2y ago
Just guessing here, but I think the name "sonnet" is the architecture, the number is the training structure / method, and the model date (not shown) is the data? So presumably with just better data they improved things signif
311.
by afro88 2y ago
You could have a second automatic step that searches the web for the title of the link and validates it or gets the correct one. Cool project!
312.
by afro88 2y ago
How do you generate / validate the links to learn more? If they're generated by the LLM there's a really high chance they are hallucinated and won't work.
313.
by afro88 2y ago
Looks like it uses bundles, which handle incremental changes: https://github.com/awslabs/git-remote-s3?tab=readme-ov-file#...
314.
by afro88 2y ago
Looks like it uses bundles rather than raw files: https://github.com/awslabs/git-remote-s3?tab=readme-ov-file#...
315.
by afro88 2y ago
Can you list some real temporary automation needs you've fulfilled? The demo shows asking for facts about space. Lower param models seem to be not great as raw chat models, so I'm interested in what they are doing well for you in
316.
by afro88 2y ago
You're right, but it's a bit of an uncharitable take on the post. Nowhere does the it say that medical intervention is unnecessary for people with conditions requiring it. The title, and quoted passage, are fully applicable to tho
317.
by afro88 2y ago
Being vulnerable only builds trauma if you're being vulnerable to damage that you aren't prepared to handle. Otherwise it's "just" adversity that you can overcome with support from close friends, family and/or
318.
by afro88 2y ago
> Be resilient. Be able to bounce back from pain or discomfort. That's kinda like saying "be rich". Resilience takes a lot of time and effort, and only really builds after going through adversity, with healing and processi
319.
by afro88 2y ago
For the use case that this project is for?
320.
by afro88 2y ago
> It is, I will not mince words, IMPOSSIBLE to build an AI detector that works longterm Like pretty much any tool involving detection of / protection from erroneous things, it's forever a cat and mouse game. There will always
321.
by afro88 2y ago
I guess define "faster". If it means getting to product/market fit earlier, agile is your guy. And that could mean eliminating waste so much that it's just 2 devs hacking on an mvp, testing it with real users every few d
322.
by afro88 2y ago
And even then it doesn't work 100%
323.
by afro88 2y ago
Consider that this work may contribute to us being able to take better care of nature. Live our lives with a lower negative impact. Sometimes study or work tangential to a goal leads to greater discovery in service to the goal.
324.
by afro88 2y ago
Does this effect last after they stop playing tetris, or just while they play? I know when I played tetris a lot, my brain seemed to be stuck on it in a way. I would close my eyes and almost see tetris shapes. I'd have dreams about it.
325.
by afro88 2y ago
You need to spend more time getting LLMs to write documentation. Without examples of how you want it to do it, it defaults to word salad that sounds impressive on the surface, but doesn't really say anything. And it very commonly uses
326.
by afro88 2y ago
Basic markdown formatting, no. But using bullet points with bold starting words, after a really waffly introduction makes it more likely they just asked an llm to write it.
327.
by afro88 2y ago
It reads like an LLM wrote it. Word salad that waffles on without any substance. In fact I think an LLM wrote most of the README. There are the telltale bullet points with bold starting words for example.
328.
by afro88 2y ago
https://chatgpt.com/share/66eb38c0-22cc-8004-9d29-024de2e39d...
329.
by afro88 2y ago
o1 gets this correct, through pure reasoning without a program. OP was likely using GPT-4(o|o-mini)
330.
by afro88 2y ago
I clicked on one of the examples, which was "State of GPT by Andrej Karpathy"

More ›