Peter Makowski

[Peter Makowski @peter_makowski]

@elonmusk …or downgrade depending on a use case 😔 before/after, the same prompt

[Peter Makowski @peter_makowski]

@AnnieKitzman @elonmusk I get terrible results now, too...

[Peter Makowski @peter_makowski]

@kennymakescoffe New model is significantly worse for me too. It's really bad especially when generating illustrations.

[Andrej Karpathy @karpathy]

Programming is changing so fast... I'm trying VS Code Cursor + Sonnet 3.5 instead of GitHub Copilot again and I think it's now a net win. Just empirically, over the last few days most of my "programming" is now writing English (prompting and then reviewing and editing the generated diffs), and doing a bit of "half-coding" where you write the first chunk of the code you'd like, maybe comment it a bit so the LLM knows what the plan is, and then tab tab tab through completions. Sometimes you get a 100-line diff to your code that nails it, which could have taken 10+ minutes before.

I still don't think I got sufficiently used to all the features. It's a bit like learning to code all over again but I basically can't imagine going back to "unassisted" coding at this point, which was the only possibility just ~3 years ago.

[Peter Makowski @peter_makowski]

My take on "Maximizing Coding Productivity with Large Language Models"

💡 Crafting effective prompts
🔁 Iterative loop to optimize prompts
📈 Automation scripts to scale LLM refactoring
🛠️ Integrating LLMs into your workflow

https://t.co/ai5kCRQU7N

[Benjamin De Kraker @BenjaminDEKR]

This is huge.

Recent academic paper shows that GPT-4 lost significant capability in code skills over a few months.

I love GPT-4 but it would be nice to have OpenAI address this.

They've been insisting that users are "imagining it," basically.

[Matei Zaharia @matei_zaharia]

Lots of people are wondering whether #GPT4 and #ChatGPT's performance has been changing over time, so Lingjiao Chen, @james_y_zou and I measured it. We found big changes including some large decreases in some problem-solving tasks: https://t.co/jgulqjvPAO

[Yann LeCun @ylecun]

This is huge: Llama-v2 is open source, with a license that authorizes commercial use!

This is going to change the landscape of the LLM market.
Llama-v2 is available on Microsoft Azure and will be available on AWS, Hugging Face and other providers

Pretrained and fine-tuned models are available with 7B, 13B and 70B parameters.

Llama-2 website: https://t.co/PKrrXgHdem
Llama-2 paper: https://t.co/aINNrXNhMb

A number of personalities from industry and academia have endorsed our open source approach: https://t.co/N7HwgW9Suh

[Ben Holmes @BHolmesDev]

Repeat after me:
One 👏
Component 👏
Per 👏
File 👏
IS 👏
WRONG 👏

[Calvin Hoenes @calvinhoenes]

Self-healing code GitHub action pipeline using @langchain and @OpenAI.
What an insane world we live in.

A 🧵 on how it works below 👇

[Playwright @playwrightweb]

Playwright 1.32 is out!

📣 Introducing UI Mode (preview)

UI Mode = Watch mode + Time-travel debugging

Watch the overview: https://t.co/K7H0tMj9LH

[Peter Makowski @peter_makowski]

My Top 5 Life Hacks for Better Sleep

https://t.co/CCVuZ4H2EH

[Peter Makowski @peter_makowski]

Hacktoberfest 2022: Registered 🎉

https://t.co/5x6KWQxiVg

[@[email protected] @nottrobin]

At work we have a blog guild, to try to support each other to do our personal blogging. A couple of days ago @peter_makowski listed some of our blogs on our team website. I love this initiative 🥰

https://t.co/apPZt8TFVO

[A11y @XA11y]

We've gotten a lot of feedback about how to improve the image description (or alt text) experience on Twitter. Today, we're launching 2 features to 3% of Twitter across Android, iOS, and Web: the public ALT badge and exposed image descriptions. 🧵 (1 of 6)

[Peter Makowski @peter_makowski]

VoiceOver can automatically generate image descriptions when an alt tag is missing by using on-device ML image recognition. 🤯

It's not perfect, but definitely better than having no description at all. Hopefully other screen readers will follow.

[Dan Hollick @DanHollick]

WCAG 3 will use a new color contrast method called APCA (Advanced Perceptual Contrast Algorithm).

It's a big improvement over the current system but there are a lot of changes to get your head around.

🧵

[Peter Makowski @peter_makowski]

💡 GitHub Tip
You can press 'a' on your keyboard to toggle annotations.

Helpful in case you find codecov annotations distracting when reviewing a PR.

[Cory House @housecor]

Love this approach to avoid spending too much time in meetings:

"Here's a document that should cover what you need. Let me know what's missing."

Insights from @ploeh:

[Peter Makowski @peter_makowski]

TypeScript Tip 💡

Use "as const" to convert the entire object to be type literals.

TypeScript uses primitive types in objects by default as it assumes that the properties might change values later.