Hacker News: dlojudice

New course on generative AI for behavioral science

dlojudice — Thu, 12 Mar 2026 02:07:27 +0000

Article URL: https://statmodeling.stat.columbia.edu/2026/03/10/new-course-on-generative-ai-for-behavioral-science/

Comments URL: https://news.ycombinator.com/item?id=47345422

Points: 1

# Comments: 0

New comment by dlojudice in "Company as Code"

dlojudice — Thu, 05 Feb 2026 15:27:34 +0000

As a programmer and founder, I think the idea is incredible, I would just change the understanding of "Code", given that what we've been hearing most lately is that "a markdown file is all you need".

I think it's not too far-fetched to think about standards, cultures, guardrails, compliance, etc. being documented, versioned, but more importantly, verifiable and applicable. In natural language, no code needed.

New comment by dlojudice in "Claude Code's new hidden feature: Swarms"

dlojudice — Sat, 24 Jan 2026 15:31:31 +0000

It feels like Auto-GPT, BabyAGI, and the like were simply ahead of their time

New comment by dlojudice in "Drug capable of reversing spinal cord injuries is in the clinical trial"

dlojudice — Fri, 23 Jan 2026 16:51:02 +0000

Researcher Tatiana Coelho de Sampaio, PhD professor, who discovered polylaminin, a drug that has proven capable of reversing spinal cord injuries in humans, worked in silence for 25 years alongside a team of biologists to achieve this breakthrough

Drug capable of reversing spinal cord injuries is in the clinical trial

dlojudice — Fri, 23 Jan 2026 16:51:01 +0000

Article URL: https://ensaiosclinicos.gov.br/news/546

Comments URL: https://news.ycombinator.com/item?id=46734699

Points: 2

# Comments: 1

New comment by dlojudice in "The AI wildfire is coming. it's going to be painful and healthy"

dlojudice — Sun, 07 Dec 2025 17:47:10 +0000

The text reminded me of one of Veritasium's latest videos [1] about power law, self-organized criticality, percolation, etc... and it also has a wildfire simulation

[1] https://www.youtube.com/watch?v=HBluLfX2F_k

Show HN: Cursor Cost Explorer

dlojudice — Tue, 11 Nov 2025 23:15:48 +0000

Hi HN — I built Cursor Cost Explorer after my monthly Cursor bill started creeping up and the official dashboard gave me almost no clue why.

I use a mix of models across sessions, sometimes one long thread, sometimes lots of tiny requests with different models, and one month my usage just ran wild. The Cursor dashboard showed not the patterns that matter: which models I used for which tasks, when cache wasn’t helping, or whether I’d be cheaper on a different plan.

So I exported my Cursor CSV and built a small analyzer that runs locally. The first time I loaded my CSV I could see a bit inside the data: a handful of expensive "thinking" model calls sprinkled across many tiny tasks. That single analysis pointed to a few concrete changes I could make that immediately cut my projected monthly spend.

What it does (short) - Parse Cursor CSVs and show where money is actually spent (by model, by day, by request) - Highlight model inefficiencies and cache patterns and give prioritized, actionable recommendations (plan change, model migrations, simple workflow fixes) - Runs entirely locally — open the demo in your browser or run `npx cursor-cost-explorer` on your machine and your CSV never leaves you, also great for agents.

UIs: - Web demo: `https://dalssoft.github.io/cursor_cost_explorer/` - CLI: `npx cursor-cost-explorer `

Would love feedback.

Comments URL: https://news.ycombinator.com/item?id=45894242

Points: 3

# Comments: 0

New comment by dlojudice in "How I use every Claude Code feature"

dlojudice — Sun, 02 Nov 2025 13:05:36 +0000

Cursor Composer appears to have this type of coupling and uses IDE resources better than other models on average.

New comment by dlojudice in "Figma Weave"

dlojudice — Thu, 30 Oct 2025 13:54:54 +0000

Node-based workflow for AI generation seems to be the right approach. Being able to chain different models (Flux for realism, Sora for video, etc.) and do actual editing in between steps is way more useful than single-shot prompting. The ComfyUI comparison is obvious but this looks a bit more polished. The branching/remixing workflow could be interesting for iteration. Also being able to create a entire workflow with just one prompt would be nice.

New comment by dlojudice in "Composer: Building a fast frontier model with RL"

dlojudice — Wed, 29 Oct 2025 23:59:07 +0000

Congratulations on your work. I spent the day working with a mix of the Composer/Sonnet 4.5/Gemini 2.5 Pro models. In terms of quality, the Composer seems to perform well compared to the others. I have no complaints so far. I'm still using Claude for planning/starting a task, but the Composer performed very well in execution. What I've really enjoyed is the speed. I had already tested other fast models, but with poor quality. Composer is the first one that combines speed and quality, and the experience has been very enjoyable to work with.

New comment by dlojudice in "Apple M5 chip"

dlojudice — Wed, 15 Oct 2025 13:45:13 +0000

Good point. Many people (including me) switched to Apple Silicon with the hope (or promise?) of having just one computer for work and leisure, given the potential of the new architecture. That didn't happen, or only partially, which is the same.

In my case, for software development, I'd be happy with an entry-level MacBook Air (now with a minimum of 16GB) for $999.

New comment by dlojudice in "Tauri binding for Python through Pyo3"

dlojudice — Mon, 13 Oct 2025 13:59:12 +0000

Going further, what can I build using it? Basically, can I use Python on a Tauri project or can I use Tauri on a Python project?

New comment by dlojudice in "Python 3.14.0 is now available"

dlojudice — Tue, 07 Oct 2025 19:34:18 +0000

+ experimental JIT compiler

this could be the beginning of something very promising

New comment by dlojudice in "Announcing Tinker"

dlojudice — Wed, 01 Oct 2025 20:41:01 +0000

> Tinker is a flexible API for efficiently fine-tuning open source models with LoRA.

It would be great if they offered inference from the trained model as well. Ideally pay per token.

New comment by dlojudice in "Cerebras systems raises $1.1B Series G"

dlojudice — Tue, 30 Sep 2025 19:28:04 +0000

OpenRouter should be responsible for this quality control, right? It seems to me to be the right player in the chain with the duties and scale to do so.

We Asked Audio Pros to Blind Test Headphones – Soundcore Won

dlojudice — Tue, 30 Sep 2025 12:10:47 +0000

Article URL: https://www.wired.com/story/we-asked-audio-pros-to-blind-test-headphones-the-results-were-surprising/

Comments URL: https://news.ycombinator.com/item?id=45424468

Points: 2

# Comments: 0

New comment by dlojudice in "ChatGPT Pulse"

dlojudice — Thu, 25 Sep 2025 17:31:15 +0000

I see some pessimism in the comments here but honestly, this kind of product is something that would make me pay for ChatGPT again (I already pay for Claude, Gemini, Cursor, Perplexity, etc.). At the risk of lock-in, a truly useful assistant is something I welcome, and I even find it strange that it didn't appear sooner.

New comment by dlojudice in "This month in Servo: variable fonts, network tools, SVG"

dlojudice — Thu, 25 Sep 2025 16:29:53 +0000

I envision the day when I'll use a secure, high-performance browser written in Rust and fully open source. That day will come...

New comment by dlojudice in "Atlassian acquires DX, a developer productivity platform for $1B"

dlojudice — Thu, 18 Sep 2025 14:38:51 +0000

Wow! $1B for a bootstrapped company [1]!! Congrats!!

[1]https://www.linkedin.com/posts/gergelyorosz_im-excited-to-sh...

New comment by dlojudice in "Tau² benchmark: How a prompt rewrite boosted GPT-5-mini by 22%"

dlojudice — Wed, 17 Sep 2025 14:06:07 +0000

I wish they had published what prompt was given to Claude to improve GPT-5-mini's performance, as well as a before and after comparison of a prompt that underwent this transformation.