Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
paradite
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
18 ms
·
211.
▲
by
paradite
2y ago
I wrote a practical guide on how to train nanoGPT from scratch on Azure a while ago. It's pretty hands-on and easy to follow: https://16x.engineer/2023/12/29/nanoGPT-azure-T4-ubuntu-guid...
212.
▲
by
paradite
2y ago
I'm already doing AI search optimizations and the early results are quite promising! https://www.reddit.com/r/ClaudeAI/s/e5tOt2MRPC
213.
▲
by
paradite
2y ago
I've replaced 99% of Google search with ChstGPT and more recently Claude. The experience is much superior. No noise, just the information that I needed.
214.
▲
by
paradite
2y ago
Ok can we reduce sudoku to something Turing complete and hence prove TypeScript is Turing complete?
215.
▲
by
paradite
2y ago
For a different user experience from the tools in the list, I am building 16x Prompt which is a standalone desktop app aimed at streamlining semi-automated AI coding workflow. It is not as well known as the others, but I do have close to 20
216.
▲
by
paradite
2y ago
It can get a little tedious if you are just using ChatGPT or Claude as it is. Also you are limited by lack of context on existing codebase. That's why there are a lot of tools that help to setup a proper workflow around these LLMs. For
217.
▲
by
paradite
2y ago
The reason is simple: You write Java. People using VS Code are writing JavaScript/TypeScript/Python that doesn't need full IDE features. And VS Code is much faster than Eclipse / IDEA kinds.
218.
▲
by
paradite
2y ago
Ops. Not sure how I missed that.
219.
▲
by
paradite
2y ago
Unlike Anthropic, OpenAI models don't have a `max_tokens` setting for API calls, so I assume the max token output limit is automatically applied to API calls. Otherwise the max token output limit stated on the models page would be mean
220.
▲
by
paradite
2y ago
Really important update that was not mentioned: gpt-4o-2024-08-06 has 16,384 tokens output limit instead of 4,096 tokens. https://platform.openai.com/docs/models/gpt-4o We don't need the GPT-4o Long Output an
221.
▲
by
paradite
2y ago
The difference is you are an individual and Delta is a public company. For enterprise purchases, you would have SLAs and contract terms to protect your company when something bad happens.
222.
▲
by
paradite
2y ago
Maybe as an operating system that Delta purchased, its kernel should not crash when a 3rd party software receives a faulty update?
223.
▲
by
paradite
2y ago
I personally think a GUI tool would provide better user experience since you can see the files, pick which ones to include as context, and check the prompt generated in real-time. So I built my own GUI version of this (plus a lot of other q
224.
▲
by
paradite
2y ago
After some initial trial and error, I've came up with a pretty good custom instruction that has been working well for over a year (for both ChatGPT and Claude): Instructions for the output format: - Output code without descriptio
225.
▲
by
paradite
2y ago
You can give 16x Prompt a try. It's GUI desktop app designed for AI coding workflow. It also doesn't automatically make code changes. https://prompt.16x.engineer/
226.
▲
by
paradite
2y ago
I think we need to the set the bar for historic moment very high for HN to not get flooded. Trump won election: yes. Trump survives attempted assassination: no. Obviously it would be historic if the outcome of assassination is different.
227.
▲
by
paradite
2y ago
Where I live, they moved to Whatsapp (dentist) and dedicated app (public hospitals) for messaging and notification.
228.
▲
by
paradite
2y ago
Assuming OpenAI wants the tech to control desktop applications using AI, the tech should be able to automate browser operations as well.
229.
▲
by
paradite
2y ago
OpenAI is on track to shipping that: https://multi.app/blog/multi-is-joining-openai
230.
▲
Coding Projects Developed Using ChatGPT
(prompt.16x.engineer)
1 points
by
paradite
2y ago
|
0 comments
231.
▲
by
paradite
2y ago
Here is mine that works pretty well on coding tasks: Instructions for the output format: - Output code without descriptions, unless it is important. - Minimize prose, comments and empty lines. - Only show the relevant code that ne
232.
▲
by
paradite
2y ago
If you buy or still have the perpetual license of Lightroom, can you use it to process raws from newly released cameras?
233.
▲
by
paradite
2y ago
Microsoft is probably one of the few companies that can sue Amazon without worrying about retaliation from Amazon. For example, GitLab would need to think twice before suing because they offer deployment on AWS.
234.
▲
by
paradite
2y ago
If you are worried about copyright, code leakage, trade secrets, IP infringement, etc, I wrote a more detailed post: https://prompt.16x.engineer/blog/ai-coding-enterprise-adopti...
235.
▲
by
paradite
2y ago
Exactly.
236.
▲
by
paradite
2y ago
I actually built a purpose-made UI (desktop app) for coding using ChatGPT, because I found that chat interface is not ideal for daily coding tasks. Curious what you think about it: https://prompt.16x.engineer/
237.
▲
by
paradite
2y ago
Update: added result section to my post with more details: https://16x.engineer/2023/12/29/nanoGPT-azure-T4-ubuntu-guid... I don't think it can go any smaller and demonstrate its capabilities. Even GPT-2
238.
▲
by
paradite
2y ago
It takes a bit of effort to setup and train a proper GPT-2 model, especially if you are not familiar with GPU drivers and python environment. Also don't try to train GPT-2 on your own machine as it takes days, even with a good gaming G
239.
▲
by
paradite
2y ago
Slightly off-topic, "open up the kimono" sounds disturbing and creepy to me as an Asian. I suspect I'm not alone in this.
240.
▲
by
paradite
2y ago
Maybe people are expecting new Macbooks? Though Apple don't usually release hardware for WWDC.
More ›