selalipop

9 ms·

31.
by selalipop 3y ago
Your reply is not indicative of someone capable of a good faith conversation on the topic, but I'll bite. I think you don't understand what the hard and easy problems are that underly the solutions I'm talking about. For exam
32.
by selalipop 3y ago
That JSON isn't something you'd type, it's something that you can programmatically generate if you have a Home Assistant setup. With super primitive wake word detection and transcription, the most you get is: - What the user
33.
by selalipop 3y ago
I think you fundamentally don't understand the topic if you're talking about two pages of text? The end user would never type in a word of that: they'd say "[Wake word] play me some music" A piece of software runnin
34.
by selalipop 3y ago
What you're describing as pre-programming is a little misleading if the "pre-programming" doesn't need to change for each specific request: a real product would provide that "pre-programming" for the user. Prom
35.
by selalipop 3y ago
> So, it will have to learn, and LLMs aren't good at learning LLMs are bad at human-like learning, but their zero-shot performance + semantic search more than make up for it. If you give an LLM access to your Spotify account via an
36.
Web Design Mistakes of 1999 (1999)(nngroup.com)
82 points by selalipop 3y ago|15 comments
37.
by selalipop 3y ago
I made a website for product ideation with AI ( https://notionsmith.ai/ ), and one of the features is essentially treating the conversation as a Turing test so you get honest feedback on your ideas rather than the default sac
38.
by selalipop 3y ago
I made https://notionsmith.ai for probing ideas like this Funnily enough when asked about itself, the generated users love the idea but wouldn't pay for it (as-is)
39.
by selalipop 3y ago
I built notionsmith.ai originally with this specific usecase in mind, but found people struggle to trust AI derived insights
40.
by selalipop 3y ago
Glad someone liked it! :) It's currently ~$100 a month, I'm working on a paid version but a lot of my users ended up being people BRIC countries so I don't mind footing the bill if they find it useful
41.
by selalipop 3y ago
I always like to say something meaningful before plugging my project, but that's exactly why I started working on https://notionsmith.ai Overall I'm really excited about LLMs being integrated deeper into product develo
42.
by selalipop 3y ago
> getting AI to be able to replicate the Eureka moments in the shower where latent information is reconstructed in parallel to processing tangential topics. I've been playing with this part specifically and it's really amazing
43.
by selalipop 3y ago
I didn't say anything about existential risk, and I'm going to assume you meant LLM since training a NN to solve sudoku puzzles has been something you could do as an into to ML project going years back: https://arxiv.or
44.
by selalipop 3y ago
This is what I mean when I say "inverse-anthropomorphization" crowd is increasingly emotion over facts. My reply to you was predicated on a compilation of centuries of scientific study on the subject of creativity. Your knee-jerk
45.
by selalipop 3y ago
Yup, it's a fun side project so I decided from the get-go I wasn't going to cater to anything non-standard It relies on WebSockets, Js, and a reasonably stable connection to run since it's built on Blazor
46.
by selalipop 3y ago
What about those of us who do understand them and just don't agree? After all you could simplify it to a layperson as: 'the LLM is just doing fancy autocomplete based on how stuff appeared in the training data so that means they&#
47.
by selalipop 3y ago
Can you try notionsmith.ai and let me know what you think? I've been working on LLMs for creative tasks and believe a mix of chain of thought and injecting stochasticity (like instructing the LLM to use certain random letters pulled fr
48.
by selalipop 3y ago
If you're manually interacting with the model, GPT 4 is almost always going to be better. Where 3.5 excels is with programmatic access. You can ask it for 2x as much text between setup so the end result is well formed and still get a r
49.
by selalipop 3y ago
If GPT 4 is working for you I wouldn't necessarily bother with this, but this is a great example of where you can sometimes take advantage of how much cheaper 3.5 is to burn some tokens and get a better output. For example I'd try
50.
by selalipop 3y ago
It depends on the domain, but chain of thought can get 3.5 to be extremely reliable, and especially with the new 16k variant I built notionsmith.ai on 3.5: for some time I experimented with GPT 4 but the result was significantly worse to us
51.
by selalipop 3y ago
Maybe the term "engineer" did the concept a disservice, but prompt engineering has a lot of parallels to the field of UX. Currently there's a lot intuition involved so it can come across as made up, but there are novel concep
52.
by selalipop 3y ago
I worked on something very much in this vein (notionsmith.ai) and feel like I should do a write up after reading this! I think a lot of people are learning these lessons in isolation, I do wish there was a centralized place where people wor
53.
by selalipop 3y ago
I don't write Rust but does this look (more) correct? extern crate winapi; use std::ptr::null_mut; use winapi::shared::guiddef::{CLSID, IID}; use winapi::um::combaseapi::{CoInitialize, CoCreateInstance}; use win
54.
by selalipop 3y ago
I've gotten them to be creative in ways that were unexpected, like outputting tokens that have absolutely no (obvious) relation to the current conversation While trying to iterate on a name for a product, I had it output an internal mo
55.
by selalipop 3y ago
I hear this often, but you don't need to surface the output of an LLM as chat I made https://notionsmith.ai and technically it's "chat based", but the user isn't exposed to 99% of the chatting: the only
56.
by selalipop 3y ago
Those are all great suggestions! I've been planning an improvement to the randomness by charting out the main themes across feedback that should be represented so there's less overlap. Even getting random names was a little tediou
57.
Show HN: I built a website that generates fake users for an idea(notionsmith.ai)
3 points by selalipop 3y ago|2 comments
58.
by selalipop 3y ago
Consumers care about where the LLM is run because of the UX: People hate how current assistants randomly break because you have a spotty signal, or take forever to respond because it's making a round trip across the internet to actuall
59.
by selalipop 3y ago
This is an interesting problem, I might start by having ChatGPT try to infer a complete list of tools and skills each position involved along with a confidence score (even those you didn't originally tell it about) From there you could
60.
by selalipop 3y ago
It actually came of my own use of the default ChatGPT interface: I was working on an indie game in my spare time and using it to spitball new mechanics with personas But it was really tedious to prompt ChatGPT into being properly critical a

More ›