Naresh Nayak Most people consider Python to be the better language to start with today. I agree -- but might suggest JavaScript as an equally good candidate. As the only language for the browser, it'll stay popular for long, and you don't need to install anything beyond the browser to learn / run it. Plus, sharing code they create is natural and easy.

Jayanth Mysore That was a camel. In fact, it was called the Camel book 🙂

In 1994, I learnt #Perl and it was fantastic. I used it to:

- Build the (unofficial) email system at Indian Institute of Technology, Madras
- Create my blog
- Author my 2nd year thesis
- Analyze bank transactions
- Transliterate languages
- Build a music search engine
- ... and much more.

In 2008, I switched to #Python - simply for cost. Google's AppEngine offered free hosting on Python. I learnt it and soon switched over.

Today, I learnt that that The Perl and Raku Conference in 2024 had fewer than 100 attendees. The next conference is moving to a smaller hotel to break-even even if only 20 people attend.

This is the language that taught me all I know about systems programming, web development, networking, functional programming, Windows programming, and much more.

Bless you, Perl, and thank you!

https://lnkd.in/gPCVAuDS

Vivek Chatrath We don't follow each other on LinkedIn. Yet. Thank God for small mercies 🙂

Anithasri Manimaran Oh, now I have not just a Bachelor's but also a Master's degree in Computer Science. And I'm a lot younger.

I'm beginning to like AI more and more by the day 🙂

Looks like XML tags are the best way to structure prompts and separate sections for an #LLM. It's the only format that all of Anthropic, Google, and OpenAI LLMs encourage.

For example:

...
...
...
...

Anthropic Docs: https://lnkd.in/gjqHGizs
OpenAI Docs: https://lnkd.in/gyyHhZCG
Google Docs: https://lnkd.in/gHuCGQJH

Alternatives are using JSON, Markdown, templating formats like Mustache/Jinja, etc.

Even Llama's system tokens seem a little XML-like.
https://lnkd.in/g6gwxj5s

Personally, I've been using Markdown so far. But it's time to switch over. (Only on the prompt side. On the generation side, Markdown still seems the best.)

Sriram Pavan well, now I know a new founder of Gramener - and a new wife I've heard of for the first time 🤣

Today, I learned that I began my career at TCS not IBM, and I never worked at the Boston Consulting Group (BCG)

I am very curious (but a bit scared) to ask an #LLM whom I'm married to.

What are the most importantly things that people need to understand in order to effectively interact with LLM-based systems like ChatGPT or Claude?

Simon Willison asked this question on Twitter and got over 150 tips.

Here's a summary:

1. Provide clear context and avoid leading questions. (Leading questions give biased answers)
2. Iterate and simplify for optimal LLM performance
3. Craft effective prompts for consistent results
4. Don’t expect human-like understanding from LLMs
5. Treat LLMs as guided children, not mind readers

... and more. Here's my categorized list from the original tweet.

https://lnkd.in/gzq_Egbp

Fascinating!

I asked 15 models the same question: 'How many letters "q" in a word "plague"' with the same system prompt: 'You are a helpful assistant'