βœ‰οΈ Not subscribed yet? Subscribe to the Newsletter

Future of Coding Weekly 2023/02 Week 1

2023-02-05 23:58

πŸ’‘ Prompting Is Programming πŸ€– Open Assistant 🐸 London Creative Coding 🌯 Type Inference & Effect Handlers

Our Work

πŸŽ₯🌯 30/01/2023, Effect Handlers via Peter Saxton

🧡 conversation

Vimeo Thumbnail

I've implemented Effect handlers in my language

This is now the whole feature set I wanted to version 2 of my language. Now all that's left is to actually use it

πŸŽ₯ Nodysseus sync via Ulysses Popple

🧡 conversation

Youtube Thumbnail

Heya, small demo showing sync between nodysseus instances. While this is on a desktop, it also works between two different devices so long as they are online at the same time

πŸŽ₯🐸 London Creative Coding - Feb 2023 πŸŽ‰πŸŽ‰πŸŽ‰ via Lu Wilson

🧡 conversation

Hi everyone I did a talk at the Peckham Digital festival yesterday! It's about spatial programming! I come in at 42:59 :)

πŸ“πŸŒ― Type inference that sticks | Jared Forsyth.com via Jared Forsyth

🧡 conversation

I've now published my post about type inference + projectional editing πŸ™‚ thanks for the feedback (HN)

πŸ“ "Type inference that sticks" (Draft) via Jared Forsyth

🧡 conversation

Thinking Together

πŸ’¬ Steve Dekorte

🧡 conversation

Should a good tool/framework/language not only "make easy things easy, and hard things possible", but also (generally) make patterns effortless, and anti-patterns painful?

πŸ’¬ Maikel van de Lisdonk

🧡 conversation

Has anyone read "compilers principles techniques and tools" and/or "modern compiler design"? Can you recommend one them or another book about the subject? I am getting some experience building compilers currently and the last year or so, but I am still rather new to it and looking for some good reference to help me grow

🐦 Tweet from @dubroy via Mariano Guerra

🧡 conversation

🐦 Patrick Dubroy πŸ‘‰ @dubroy@hachyderm.io: What are some great ways to present code in a book/docs? Especially when the code is growing, being modified, etc.?

I like what @munificentbob does in Crafting Interpreters…what are some other good examples?

Devlog Together

πŸ’¬ Jason Morris

🧡 conversation

Spending most of this week taking tree-structured explanations, and generating something approximating a series of paragraphs, then annotating them with cross references like a graph for navigation purposes. The result is persuasive for my target audience. An explanation that they can read and understand and navigated easily is a big deal for demonstration purposes. But I am hating it, because it is technologically trivial. Any tips for getting through the slog of actually doing the thing that is time consuming, valuable, but uninspiring?

🧡 conversation

Also, if anyone has any examples of an interface that uses an auto-fill to let you select a prototype statement, then let's you fill in the blanks in that prototype, I am looking for UI inspiration.

🧡 conversation

Working on a paper for the ICLP this week on Blawx as a visual interface for logic programming. Any tips for getting accepted at ICLP would be appreciated. πŸ™‚

Content

πŸ’‘ Prompting Is Programming: A Query Language For Large Language Models via Mariano Guerra

🧡 conversation

Prompting Is Programming: A Query Language For Large Language Models

we present the novel idea of Language Model Programming (LMP). LMP generalizes language model prompting from pure text prompts to an intuitive combination of text prompting and scripting. Additionally, LMP allows constraints to be specified over the language model output. This enables easy adaption to many tasks, while abstracting language model internals and providing high-level semantics.

To enable LMP, we implement LMQL (short for Language Model Query Language), which leverages the constraints and control flow from an LMP prompt to generate an efficient inference procedure that minimizes the number of expensive calls to the underlying language model.

We show that LMQL can capture a wide range of state-of-the-art prompting methods in an intuitive way, especially facilitating interactive flows that are challenging to implement with existing high-level APIs. Our evaluation shows that we retain or increase the accuracy on several downstream tasks, while also significantly reducing the required amount of computation or cost in the case of pay-to-use APIs (13-85% cost savings).

PDF Version

πŸ“ Legends of the Ancient Web via Kartik Agaram

🧡 conversation

Maciej CegΕ‚owski gives great talks and also puts up excellent transcripts of them. For example: Legends of the Ancient Web

But just go read anything with pictures on his talks page.

πŸ€– Open Assistant via Nilesh Trivedi

🧡 conversation

This just came out

πŸ“’ Future of Coding Tools have new URLs:

Contents Β© 2024 Mariano Guerra - Powered by Nikola