GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2 (magazine.sebastianraschka.com)

I am starting a new project and I would like to experiment with some AI assistants when it comes to coding. It will be a Python project that will access Microsoft cloud using the Graph 1.0 API.

One advantage of the situation is that this project is a bit standalone and I don't have to feed it 2000 already existing classes from an existing legacy codebase to give it a context to work with.

What would you recommend me?

Are there any dark patterns involved, e.g. de-facto un-cancellable subscriptions that will make me cancel my card instead?

I would be happy hearing your experience and tales from the battlefield.

Comments (10)

mikewarot · 1h ago

I've heard, and it seems reasonable to me, that you should never get completely out of the loop when it comes to the actual code. Type it all in yourself so that you can get a real sense of it. Do not copy and paste.

If you do otherwise you're just creating legacy code at astonishing speed. This is fine as long as you throw it away after you're done.

cellis · 1h ago

Claude Code certainly not as easy to engineer with, though it is less expensive. For instance the @feature isn’t as robust as cursors ime. Also no shift+enter is quite a pain. Linting doesn’t “just work”, cursor with Claude 4.0 max is really thorough, I think even better than GPT-5. Not that Sonnet is better but that whatever “ensemble” of models cursor uses with sonnet seems to both adhere and tool call better than with GPT-5. GPT-5 often says what it will do and then says “say go and I’ll go” or says “you should run command x”, but doesn’t just DO it. Also for bug fixes in difficult codebases nothing beats Gemini 2.5 pro

the__alchemist · 1h ago

  - Don't let it create function signatures or structs/classes.
  - Instruct it explicitly not to change variable names, or things unrelated to your query
  - Pass it exactly the information and code it needs to solve the task, and no more
  - If it starts going off the rails, or you're more than a few iterations in, start over. (Likely with modified context, including any parts of its results it did correctly)

snaga · 1h ago

I have tried the spec-driven development with Gemini (Gemini-CLI) and VS Code for a few weeks.

It is definitely inspired by Kiro by Amazon. (unfortunately, I'm still on the wishlist.)

It works fine for me, and I would recommend this approach to understand how AI-assisted coding works.

PretzelPirate · 1h ago

I started with Github Copilot and VSCode. It has a free tier and is an easy way to try out AI-assisted development.

viraptor · 3h ago

Cursor is kind of the default these days and has reasonable pricing. Both the completion and the agent work well. You could nitpick various options, but really it's ok and you can do a lot worse than Cursor.

An alternative is buying credit for a specific provider and using that with Aider. Which is also not bad.

None of the major players are likely to abuse your card. Just make sure you're either using a prepaid plan or prepaid credits with no auto topups.

bad_haircut72 · 1h ago

Cursor certainly isnt the default ! No idea where you're getting that? If anything I would say Claude Code is the default but realistically all the tools are converging to be much of a muchness

the__alchemist · 1h ago

Claude is so Mid-summer 2025. Crayon is the new hotness; everyone's using it and getting 11x gains.

uberman · 3h ago

All the free tiers of the top products will allow you a days worth of coding if you are taking bite sized tasks such as defining methods and actually proofing to understand the results.

MilnerRoute · 22m ago

Lex Fridman did an interview with Ruby on Rails creator David Heinemeier Hansson. I think DHH actually keeps his AI in a separate window, and insists on retyping the suggestions it displays. (The act of typing the code somehow plants it more deeply in his memory, and he finds it more education.)

From the interview I got the impression that AI can help you learn or rob you or learning, depending on how you use it. Like, you can go fast or you can go slower-but-more-educational... Depends on what you're after, I guess.

AOL closes its dial up internet service (ispreview.co.uk)

Try and (ygdp.yale.edu)

GPT-OSS vs. Qwen3 and a detailed look how things evolved since GPT-2 (magazine.sebastianraschka.com)

Diffusion Language Models Are Super Data Learners (jinjieni.notion.site)

Show HN: Engineering.fyi – Search across tech engineering blogs in one place (engineering.fyi)

MCP: An (Accidentally) Universal Plugin System (worksonmymachine.ai)

Type (YC W23) is hiring a founding engineer to build an AI-native doc editor (ycombinator.com)

Booting 5000 Erlangs on Ampere One 192-core (underjord.io)

Writing simple tab-completions for Bash and Zsh (mill-build.org)

Inside OS/2 (1987) (gitpi.us)

Zig's Lovely Syntax (matklad.github.io)

Flintlock – Create and manage the lifecycle of MicroVMs, backed by containerd (github.com)

Abogen – Generate audiobooks from EPUBs, PDFs and text (github.com)

Open Lovable (github.com)

How I code with AI on a budget/free (wuu73.org)

GPT-5: It Just Does Stuff (oneusefulthing.org)

Curious about the training data of OpenAI's new GPT-OSS models? I was too (twitter.com)

Show HN: The current sky at your approximate location, as a CSS gradient (sky.dlazaro.ca)

Abusing Entra OAuth for fun and access to internal Microsoft applications (research.eye.security)

My Lethal Trifecta talk at the Bay Area AI Security Meetup (simonwillison.net)

OpenFreeMap survived 100k requests per second (blog.hyperknot.com)

How Potatoes Evolved (nhm.ac.uk)

POML: Prompt Orchestration Markup Language (github.com)

The Framework Desktop is a beast (world.hey.com)

Sunlight-activated material turns PFAS in water into harmless fluoride (phys.org)

“The Hollow Men” at 100 (prufrock.substack.com)

A CT scanner reveals surprises inside the 386 processor's ceramic package (righto.com)

The current state of LLM-driven development (blog.tolki.dev)

Melonking Website (melonking.net)

LHC's New Chip Tackles Radiation Challenges (spectrum.ieee.org)

Ch.at – A lightweight LLM chat service accessible through HTTP, SSH, DNS and API (ch.at)

Quickshell – building blocks for your desktop (quickshell.org)

Don't “let it crash”, let it heal (zachdaniel.dev)

Long-term exposure to outdoor air pollution linked to increased risk of dementia (cam.ac.uk)

Did California's fast food minimum wage reduce employment? (nber.org)

Debian 13 “Trixie” (debian.org)

People returned to live in Pompeii's ruins, archaeologists say (bbc.com)

LLMs Aren't World Models (yosefk.com)

An engineer's perspective on hiring (jyn.dev)

Who got arrested in the raid on the XSS crime forum? (krebsonsecurity.com)

I want everything local – Building my offline AI workspace (instavm.io)

P-fast trie, but smaller (dotat.at)

An AI-first program synthesis framework built around a new programming language (queue.acm.org)

Employees spotting problems help the business, but leaders empower flatterers (phys.org)

ESP32 Bus Pirate 0.5 – A hardware hacking tool that speaks every protocol (github.com)

The importance of offtopic (blog.tadzik.net)

A Simple CPU on the Game of Life (2021) (nicholas.carlini.com)

GPTs and Feeling Left Behind (whynothugo.nl)

R0ML's Ratio (blog.glyph.im)

MCP overlooks hard-won lessons from distributed systems (julsimon.medium.com)

Ask HN: Advice for someone who wants to try AI-assisted coding?

Comments (10)