Hacker News: ammar_x

New comment by ammar_x in "DeepSeek makes the V4 Pro price discount permanent"

ammar_x — Fri, 22 May 2026 18:25:23 +0000

You can use V4 Pro with Claude Code [1].

I tried it and it's impressive.

[1]: https://api-docs.deepseek.com/quick_start/agent_integrations...

New comment by ammar_x in "My trick for getting consistent classification from LLMs"

ammar_x — Tue, 21 Oct 2025 18:00:45 +0000

My "trick" was to divide things into batches (which can be big with LLMs with larger context sizes) and classify the items in each batch, then take the resulting categories from each batch and feed them into an LLM to group semantically similar categories into groups with a representative category for each group. The representative category can be chosen from the group or created by the LLM. This is an over-simplification of the process but that's the gist of it.

New comment by ammar_x in "DeepSeek OCR"

ammar_x — Mon, 20 Oct 2025 09:44:10 +0000

Language support is not mentioned in the repo. But from the paper, it offers extensive multilingual support (nearly 100 languages) which is good, but I need to test it to see how it compares to Gemini and Mistral OCR.

New comment by ammar_x in "Claude Skills"

ammar_x — Fri, 17 Oct 2025 16:50:50 +0000

Claude Skills seem to be the option that offers highest flexibility to add more capabilities at most simplicity. Better than MCP in my opinion. Hope it becomes a standard and get adopted by OpenAI and the rest of labs.

New comment by ammar_x in "Exploring Goodreads data: Analysis of 10M books"

ammar_x — Tue, 13 Aug 2024 22:23:24 +0000

Good question! I selected the edition with the smallest Goodreads ID¹ that has the publication date and cover photo available. If all editions don't have publication date nor cover photo, then we get the one with the smallest ID.

And you're right, in a few cases, this resulted in getting less widely read editions for some books.

1: Assuming smaller ID means earlier addition to Goodreads' database.

Exploring Goodreads data: Analysis of 10M books

ammar_x — Tue, 13 Aug 2024 16:59:19 +0000

Article URL: https://ammar-alyousfi.com/2024/exploring-goodreads-data-an-analysis-of-10-million-books

Comments URL: https://news.ycombinator.com/item?id=41237194

Points: 4

# Comments: 2

New comment by ammar_x in "FastHTML – Modern web applications in pure Python"

ammar_x — Tue, 30 Jul 2024 13:35:46 +0000

Hi Jeremy, congratulations for the launch.

How does this compare to Dash?

I've used Dash for many applications, so I'm wondering what are the advantages of FastHTML?

New comment by ammar_x in "Ask HN: What are your favourite websites that display a lot of data / tables?"

ammar_x — Sun, 26 May 2024 18:00:00 +0000

Been looking for such a website to show weather for the whole year like this. Thanks for sharing.

New comment by ammar_x in "Ask HN: Most successful example using LLMs in daily work/life?"

ammar_x — Mon, 20 May 2024 21:43:38 +0000

I have Raycast extensions for GPT and Claude models. Whenever I have a question, the most powerful LLMs in the world are two key strokes away.

This way is easier than going to the browser then ChatGPT tab for example then creating a new chat.

I found myself using LLMs more and getting more out of them because of this frictionless interaction. They've become more of actual "helpful assistants."

New comment by ammar_x in "GPT-4o's Memory Breakthrough – Needle in a Needlestack"

ammar_x — Wed, 15 May 2024 12:28:13 +0000

The article compares GPT-4o to Sonnet from Anthropic. I'm wondering how Opus would perform at this test?

New comment by ammar_x in "Reflex – Web apps in pure Python"

ammar_x — Tue, 08 Aug 2023 15:58:39 +0000

How does it compare to Plotly Dash?

A study confirms: Big changes in GPT-4 performance since its launch

ammar_x — Wed, 19 Jul 2023 11:09:45 +0000

Article URL: https://twitter.com/emollick/status/1681482007140761601

Comments URL: https://news.ycombinator.com/item?id=36784554

Points: 2

# Comments: 0

What's your favorite interface for GPT API?

ammar_x — Sun, 25 Jun 2023 11:17:04 +0000

When you use OpenAI GPT API, you want an interface where you can quickly change model used, temperature, etc. and get results with nice formatting.

There might be extra features like prompt library/templates, text to speech, etc. Also, ways to send long chat history to the API without exceeding the context limit.

So what's you favorite? Free or paid.

For me, I found Chatbox [1] to be good for now. I tried ChatGPT Next Web [2] which seems to have advanced features but I couldn't learn how they work because most of the documentation is in Chinese. I tried Lore [3] a while ago and it seemed to have a great UX but when I wanted to try it again, I found that my trial was over and couldn't try it more nor know if it's still being updated.

[1] https://github.com/Bin-Huang/chatbox

[2] https://github.com/Yidadaa/ChatGPT-Next-Web/

[3] https://thellm.app/

Comments URL: https://news.ycombinator.com/item?id=36467024

Points: 1

# Comments: 0

New comment by ammar_x in "Ask HN: What's your go-to platform for written discussions, and why?"

ammar_x — Wed, 21 Jun 2023 13:59:48 +0000

Can you explain more? Like which tool do you use for this wiki page? Or is it an internal tool? And do you use it to write meeting notes and then discuss on the same page?

Ask HN: What's your go-to platform for written discussions, and why?

ammar_x — Wed, 21 Jun 2023 13:28:51 +0000

Writing is still the most effective way for long-form discussions in my opinion to reach the best decision.

I find that tools like Slack aren't quite optimized for this.

What tools does your organization utilize for such in-depth written discussions?

Comments URL: https://news.ycombinator.com/item?id=36418184

Points: 10

# Comments: 4

New comment by ammar_x in "React, but in Python"

ammar_x — Mon, 12 Jun 2023 11:20:41 +0000

How is this different or better than Dash or Streamlit?

New comment by ammar_x in "Ask HN: Is it just me or GPT-4's quality has significantly deteriorated lately?"

ammar_x — Mon, 05 Jun 2023 14:11:39 +0000

Like many people here have noticed, it's definitely less quality now than before. It's annoying to be honest to reduce the quality significantly without a notice while we are paying the same amount. I'm willing to pay $40 for the original GPT-4, though.

GPT-4 was remarkable 2 months ago. It could handle complex coding tasks and the reasoning was great. Now, it feels like GPT-3. It's sad. I had many things in mind that could have been done with the original GPT-4.

I hope we see real competitors to GPT-4 in terms of coding abilities and reasoning. Absence of real competitors made it a reasonable option for "Open"AI to lobotomize GPT-4 without a notice.

New comment by ammar_x in "Ask HN: What's your favorite GPT powered tool?"

ammar_x — Wed, 17 May 2023 07:58:28 +0000

I couldn't find it on the app store, can you post the link?

Ask HN: Have you noticed decreased quality in GPT-4 reasoning recently?

ammar_x — Tue, 16 May 2023 08:03:01 +0000

I noticed a significant decrease in the model reasoning abilities in the last few days, especially when it comes to writing code. Has anyone else experienced this?

Given that these models are being used in all sorts of applications all over the world, I think it's concerning that OpenAI can lobotomize a model in secret and no one can verify that and do something about it.

Comments URL: https://news.ycombinator.com/item?id=35958878

Points: 4

# Comments: 5

Ask HN: What is the main communication channel at your remote company?

ammar_x — Thu, 16 Mar 2023 18:02:56 +0000

Especially if you are at a remote-first company. What is the main communication tool/channel?

Is it Slack? meetings? email? Asana? a combination of these? another innovative way? How do you use your tools to manage discussions and communication?

I'm sick of using Slack as the main communication channel at my company. Many discussions get lost as new discussions appear. Threads are awful but sometimes necessary. Link previews are distracting. Etc. I feel we are misusing it.

We do meetings and emails of course, and Asana, but Slack is the main and most used thing.

I'm trying to find a way that's good for work-related discussions and at the same time provides some social element (we are fully remote, we need to feel the company is alive)

So, please share what you have!

Comments URL: https://news.ycombinator.com/item?id=35185741

Points: 1

# Comments: 0