Hacker News Reader

tydunnon Show HN: LLM-Tetris-A Classic Tetris Game in Go, 95% AI-Generated

Super cool! I enjoyed reading Harper Reed's post too. For anyone, who hasn't seen it: https://harper.blog/2025/02/16/my-llm-codegen-workflow-atm/

tydunnon Ollama is now available on Windows in preview

I use Continue.dev's new tab-autocomplete [1] (disclaimer: I am one of the authors of this open-source project)

[1] https://continue.dev/docs/walkthroughs/tab-autocomplete

tydunnon I worry our Copilot is leaving some passengers behind

> Copilot is encouraging us to block users unnecessarily, by suggesting obviously flawed code, which is wrong on every level: wrong ethically, wrong legally, and the wrong way to build software.

I share many of the same worries as the author. This is why I think teams need to build and run their own Copilot-like systems, so that they can guide the suggestions they receive. Each developer and team has their own way of building software, and they need to be able to shape and evolve the suggestions they receive to fit their definition of the "right" way: https://blog.continue.dev/its-time-to-collect-data-on-how-yo...

tydunnon Ollama is now available on Windows in preview

I've been playing around with it for the last couple days on my Windows machine, using it for local tab-autocomplete in VS Code, and it's been just as good as it is on my Mac

tydunnon Ask HN: Who is hiring? (February 2024)

At Continue, we are on a mission to make building software feel like making music. We are creating the open-source autopilot for VS Code and JetBrains——the easiest way to code with any LLM (https://github.com/continuedev/continue).

You are likely a good fit if you

- have founded or want to found your own startup one day

- have experience with frontend, backend, ML technologies

- are enthusiastic about AI/LLMs, open source, developer tools

- get excited about supporting users and helping customers

- want to work in-person in SF the majority of the time

More info: https://www.ycombinator.com/companies/continue/jobs/smcxRnM-...

tydunnon Ask HN: Open source LLM-powered SQL tools?

I have not used them, but the SQLCoder models look promising: https://github.com/defog-ai/sqlcoder

tydunnon Ask HN: Who is hiring? (December 2023)

At Continue, we are on a mission to make building software feel like making music. We are creating the open-source autopilot for software development—an IDE extension that brings the power of ChatGPT to VS Code and JetBrains (https://github.com/continuedev/continue).

You are likely a good fit if you

- have founded or want to found your own startup one day

- have experience with frontend, backend, and ML technologies

- are enthusiastic about AI/LLMs, open source, developer tools

- get excited about supporting users and helping customers

- want to work in-person in SF the majority of the time

More info: https://www.ycombinator.com/companies/continue/jobs/smcxRnM-...

tydunnon Generate and Pray: Using SALLMs to Evaluate the Security of LLM Generated Code

> RQ2 Findings: StarCoder generated more secure code than CodeGen-2B, CodeGen-2.5 7B, GPT-3.5 and GPT-4.

I wonder why StarCoder might generate more secure code

tydunnon One LLM won't rule them all

Reminded me of "Does One Large Model Rule Them All?" from earlier this year: https://maithraraghu.com/blog/2023/does-one-model-rule-them-...

tydunnon Language gaps constrain generative AI development

I wonder if this is the same for programming languages too

tydunnon Ask HN: Who is hiring? (November 2023)

You are likely a good fit if you

- have founded or want to found your own startup one day

- have experience with frontend, backend, and ML technologies

- are enthusiastic about AI/LLMs, open-source, developer tools

- get excited about supporting users and helping customers

- want to work in-person in SF the majority of the time

More info: https://www.ycombinator.com/companies/continue/jobs/smcxRnM-...

tydunnon Phind Model beats GPT-4 at coding, with GPT-3.5 speed and 16k context

This is awesome. Are you planning to open-source the V7 model?

tydunnon RedPajama v2 Open Dataset with 30T Tokens for Training LLMs

This is a lot of tokens. Llama 2 was trained on two trillion tokens [1]

[1] https://arxiv.org/abs/2307.09288

tydunnon Executive Order on the Safe, Secure, and Trustworthy Development and Use of AI

Companies developing or demonstrating an intent to develop potential dual-use foundation models to provide the Federal Government, on an ongoing basis, with information, reports, or records ... (i) any model that was trained using a quantity of computing power greater than 10^26 integer or floating-point operations, or using primarily biological sequence data and using a quantity of computing power greater than 10^23 integer or floating-point operations; and (ii) any computing cluster that has a set of machines physically co-located in a single datacenter, transitively connected by data center networking of over 100 Gbit/s, and having a theoretical maximum computing capacity of 10^20 integer or floating-point operations per second for training AI.

Training large models in the US now requires reporting

tydunnon CodeFusion: A Pre-Trained Diffusion Model for Code Generation

How have you tried to use them?

tydunnon The 'it' in AI models is the dataset

We seem to always be relearning this every few years: https://hbr.org/2021/07/ai-doesnt-have-to-be-too-complicated...

tydunnon The GitHub Recent Bugs Dataset for Evaluating LLM-Based Debugging Applications

I've been surprised that evaluation dataset leakage into training dataset is not discussed more, given that LLM creators often don't reveal their training data

tydunnon Refreshing LLMs

This is awesome. Thanks for sharing!

tydunnon Add an AI Code Copilot to your product using GPT-4

Continue (disclaimer: I am a creator) is a VS Code extension that enables you to highlight a section of code or file and tell it to comment all of your functions and ensure that the typing is correct. Here is the point in our demo video where we show this: https://youtu.be/3Ocrc-WX4iQ?t=31. You could even create a custom slash command to do this if you do it frequently and want it done in the same specific way: https://continue.dev/docs/customization#custom-commands