Week 1 Video AI Training: Watch the first two segments of the Google AI Essentials Preview Course - Introduction to Artificial Intelligence and Use AI Tools to Boost Productivity.   These two videos give you an accessible, high level overview of AI tools for productivity.

AI Training #2: Coursera Generative AI for Everyone You can register for free for this course by registering with your Google account and choosing to audit the course.   This entire course is worthwhile, but for this week we especially recommend LLMs as a thought partner, AI is a general purpose technology, Writing, Reading, and What LLMs Can and Cannot Do. 

Get Inspired: Watch Microsoft AI CEO Mustafa Suleyman at his Ted Talk in April 2024 - What is an AI Anyway?

Reading: Train Your Brain to Work Creatively with Gen AI - Like all tools, it takes practice to improve how you use them. 

AI Chatbots look simple to use, but they can be deceptively complicated.   We've linked this full long blogpost in the Go Deeper section below, but this one section is quick and may help you understand why something so simple can be so frustrating.  

Throughout BLAM, we'll be trying a variety of tools, but the most important one is CBorg.   CBorg is Berkeley Lab's multi-AI portal, where you can chat with and use the most powerful AI models all in one place - including ChatGPT, Claude, LLama, and Gemini.   

Try It #1:  Login to Cborg and experiment with using different AI models.   Does one consistently outperform another on your query?   

Try It #2: gemini.google.com is your direct access to Gemini from Google Workspace.   New features are currently rolling out to everyone at Berkeley Lab, so there may be variations in what Gemini can do - but explore how it can work with Gmail.

Lab Poetry Challenge

Large Language Models are good with words - even poetry.  For your first weekly challenge, use Cborg or Gemini to compose a poem about life at the Lab - it could be a haiku, a limerick, or any other kind of poem.   Pick something amusing, mundane, or interesting about Berkeley Lab Life.   When you've got a poem you like, learn how to use the BLAM Chat Rooms (Important!) and then share your work and the prompt you used to make it in the Week 1 Poetry Challenge thread here.   While you're there, peruse other people's submissions and upvote them by adding reactions.   The BLAM! team will be awarding bucketlist awards to popular entries and our personal favorites. Oh, and keep it work-appropriate please.

Click Here to Subscribe to the BLAM Calendar to see the latest information about all the events planned for the next six weeks.  

NOTE: All past webinar recordings are available here.

Tuesday January 28 2-3pm - For Coders, join Tim Fong as he walks you through how to setup and use Omni Engineer. Webinar.   Slides Here for Mac.  Document here for Windows.    Recording here.

Wednesday January 29 11am-Noon - Join Adam Stone for an overview of Lab AI resources from CBorg to GPUs.  From productivity tools to consulting for scientists, IT has you covered. Webinar.  Recording here.

Wednesday January 29 10.30-Noon - ScienceIT and Coders Office Hours - Stop by and get answers to your questions about developing with LLMs. Online.

Each Week, we'll feature some additional optional resources to dive deeper into AI.

Things we learned about LLMs in 2024:  Feeling like you might be behind on the AI/LLM News?  Let Simon Willison summarize all the substantive AI news of 2024 for you in one digestible blog.  

Ready for more?   Tina Huang summarizes Google's 8 Hour AI Essentials Course in 15 Minutes.   It's good too.  If you watch this, you'll have the high level outline of the next 6 weeks.

OK, bear with us because intermediate to advanced covers a lot of ground at LBL - from regular users of LLMs to computational scientists pushing their boundaries - so understand that not every 201 entry will be right for everyone.  Our goal with 201 is to give people who are already familiar with and comfortable with these tools new knowledge and skills to help improve their use.   

Get the Background: What is AI and how do LLMs fit in?

How Many Rs in Strawberry:  Why did it take so long for LLMs to know the answer and what does it mean for our understanding on their limitations. 

A good place to try advanced LLM tech is Google's AI Studio.  

Try It #1: Experiment with realtime streaming your screen to Gemini in AI studio.   Does it understand what you're looking at and can it answer questions?   This video from Tiny Technical Tutorials gives you a good introduction to what it can do.  

Try It #2: Create a Gem with gemini.google.com - Gems are customized personas with instructions and context to help you in certain tasks.  Try giving instructions, prompts, and background reading.  (Note: For some LBL users the full Gem feature may not be available until next week).  Learn more about Gems here.

Getting LLMs to make diagrams can be a challenge.  CBorg has an open source version of a tool similar to Claude Artifacts which can make the process easier - LibreChat Code Artifacts. You'll need to toggle it on under Settings/Beta Features.

Your challenge this week is to create a funny process diagram for a weird/funny factory.   Use CBorg with Artifacts to create your diagram, learn how to use the BLAM! Chat Rooms (Important!) and then  share your prompt and diagram on BLAM! Chat.  Feel free to share the link directly to your Chat in CBorg too.   

Getting the most out of LLMs as programming assistants can be challenging.   Learn about the techniques you can use to get the most out of them. 

LinkedInLearning's Pair Programming with AI provides a solid introduction to the basics of working with LLMs to code.  You may need to visit go.lbl.gov/linkedinlearning to get authenticated first. 

Here is a link to the first coder webinar featuring Tim Fong.

Turn on CBorg's Artifacts feature (see Week 1 Intermediate Challenge above) and use it to create code for a simple visualization.  For example, you could say "Create a react component for a simple bouncing ball visualization with a blue ball bouncing on a white background in a 400x400px box." and see the results in realtime.  Now try asking for changes to the code.  How does it do? 

The Coder challenge this week is to get your environment set up to use LBL Omni Engineer and Cborg.   Attend Tim Fong's Tuesday 2pm Webinar or watch the replay to learn how to get set up.  We'll be using this environment throughout the rest of BLAM!

The pace of change in AI tools to help coders has been astounding.   This blog post from Val Town chronicles the last year of developments in AI coding and their efforts to include those features in their product.  It's a good whirlwind tour of how things have evolved.