Hacker News Reader

wfhBrianon When will global warming hit the landmark 1.5 ºC limit?

Apple TV's "Extrapolations" is an interesting dramatization of this. Highly recommended.

wfhBrianon LangChain: The Missing Manual

LangChain is undeniably the best option for building demos.

wfhBrianon Apple Restricts Employee Use of ChatGPT, Joining Other Companies Wary of Leaks

It also disables their plugins (browsing, code interpreter, etc.)

wfhBrianon Google I/O and the Coming AI Battles

> ChatGPT is finished! Now that Bard has launched here 10 tips on how to..

As if the spammy AI threads couldn't get worse, they managed to do it.

wfhBrianon Evidence.dev – Business Intelligence as Code

If you like this, Dataview[1] is great for querying your markdown notes.

[1] https://blacksmithgu.github.io/obsidian-dataview/queries/str...

wfhBrianon Heavy EVs Could Collapse Old Parking Garages: Report

Similarly, why don't we pay more to drive a vehicle that is more dangerous to both pedestrians and other drivers.

wfhBrianon To supercharge learning, look to play

There are plenty of capable people, but having to take a vow of poverty understandably prevents a lot of them from becoming educators.

wfhBrianon “I've switched over from Mem to Obsidian when I found this plugin”

I've been working on this passion project of mine, designed to address my personal needs for a PKM (Personal Knowledge Management) system. To my delight, it seems that many others in the community have found it valuable as well.

wfhBrianon Show HN: Compress GPT-4 Prompts

Tried twice and didn't get usable results. The first reduced the tokens by 80% and the results were better than I would've expected.

The second try reduced the same prompt by only 20% and the results were worse than the first try.

Both missed a key detail about using a specific term in the response.

Interesting idea. Not sure how useful it will be with costs dropping fast. But if you can get it to work better, it might be useful for recursive prompt-chains like Auto-GPT or babyAGI.

wfhBrianon ‘AI Pause’ Open Letter Stokes Fear and Controversy

Can you point to any sufficient evidence that there is a real risk that 6 months could mitigate?

If there was cause for real concern, these people would be communicating there evidence instead of just stoking fears.

wfhBrianon JavaScript-algorithms: Algorithms and data structures implemented in JavaScript

I would love to see HNSW in JS.

wfhBrianon Prompt engineering DaVinci-003 on our own docs for automated support (Part I)

This is correct from what I've seen, but it's not well documented. Also, that fine-tuning is "better for training " style than knowledge."

wfhBrianon DeepMind Dramatron: a new tool for writers to co-write theatre and film scripts

This is incredibly underwhelming.

It's literally just text inputs.

Each is connected to a prompt and uses content from the previous text input to complete a the prompt for the next input. Inputs are editable at each stage. AKA Prompt chaining.

The "Settings" prompt returned nothing. And the "dialogue" prompt hung up, still saying "Generating..." for five minutes now. And there isn't any other option but to refresh the page.

Still going to check out the attached 100-page research paper, but not as urgently since testing out the app.

wfhBrianon Ask HN: How would you build a ChatGPT detector?

I came across this[1] recently. I tested it on content that I used an GPT to help write. It does a good job detecting short content that wasn't "refined" much. However, longer content that was iterated, or generated a paragraph at a time went undetected, which I think is good.

[1] https://contentatscale.ai/ai-content-detector/

wfhBrianon Ask HN: Who wants to collaborate?

Let's WFH while building the future of work. Get support building and promoting your WFH idea, contact in bio.