Author | Lightnews

Dale Lane @dalelane.co.uk · 17h

Wow, that got dark quick. My fields-and-woodlands walk home from the office has become an excitingly dark one... curse you, wonky tilted Earth! ;-)

Dale Lane @dalelane.co.uk · 1d

on my way home after an inspiring two days in Cambridge for #ClubsCon25

www.linkedin.com/posts/code-c...

#clubscon25 | Code Club

✨Code Club Conference 2025 What a fantastic day of creativity, learning, and community here in Cambridge. 💚 💡 We joined workshops on LLMs, micro:bits, robotics, and Scratch, enjoyed talks on inclusive learning, generative AI, and the Code Club journey, and were inspired by global highlights from India, Kenya, and South Africa, showcasing innovation in clubs around the world. Thank you to everyone who joined us and shared their ideas. We can’t wait for more tomorrow! #ClubsCon25

www.linkedin.com

Dale Lane @dalelane.co.uk · 2d

it was a bit ambitious to try and have a night out after being awake for 34 hours - I managed one pint then bailed before 8pm to walk back to my B&B and sleep 😴

Dale Lane @dalelane.co.uk · 2d

lol - yeah, that wasn't ideal 😂

Dale Lane @dalelane.co.uk · 2d

experimenting with #LLM context in #Scratch at #ClubsCon25

asking a language model "What is the capital city of France?" and getting the answer "Paris"

asking a language model "What is the capital city of France?" with the context "answer in the style of a pirate"

2 1

Dale Lane @dalelane.co.uk · 2d

I presented at #ClubsCon25 about introducing language models to Code Clubs by getting students to build and make stuff with an #LLM in #Scratch

It was loosely based on what I was writing about in March, but was a more practical take now that I've run this in a few classrooms

Exploring Language Models in Scratch with Machine Learning for Kids

In this post, I want to share the most recent section I've added to Machine Learning for Kids: support for generating text and an explanation of some of the ideas behind large language models. youtu...

dalelane.co.uk

Dale Lane @dalelane.co.uk · 3d

I just landed at Heathrow after a busy week in New Orleans. Now I need to get through security, get to my car, and drive to Cambridge in time for the start of #ClubsCon25

Hopefully they have a limitless supply of coffee for when I get there... I think I'll need it.

Dale Lane @dalelane.co.uk · 3d

My talk at #Current25 on Wednesday was an overview of the most common patterns for bringing #AI to #Kafka topics.

The second pattern I talked about was using #agentic AI to respond to events, or the output from event stream processing.

Triggering agentic AI from event streams

In this post, I describe how agentic AI can respond autonomously to event streams. I spoke at Current on Wednesday, about the most common patterns for how AI and ML are used with Kafka topics. I had ...

dalelane.co.uk

1

Dale Lane @dalelane.co.uk · 3d

Current is over, so I've started the journey back to the UK with a 7am bus to the airport. 🥱

I'm not heading home just yet, though - I need to be in Cambridge tomorrow morning for #ClubsCon25

codeclub.org/en/clubs-con...

Dale Lane @dalelane.co.uk · 3d

fun in the French Quarter last night - I'm guessing Halloween week is partly responsible for the outfits and energy I saw... or New Orleanians just go all in for a night out 😆

1

Dale Lane @dalelane.co.uk · 4d

sticking with my usual habit when visiting a new city for work and going for the tallest view I can find - with a drink at the top of Vue Orleans

2

Dale Lane @dalelane.co.uk · 4d

My talk at #Current25 yesterday was an overview of the most common patterns for bringing AI to #Kafka topics.

The first pattern I talked about was using AI to enhance what we can do with event stream processing.

Using AI to augment event stream processing

In this post, I describe how artificial intelligence and machine learning are used to augment event stream processing. I gave a talk at a Kafka / Flink conference yesterday about the four main patter...

dalelane.co.uk

1

Dale Lane @dalelane.co.uk · 4d

I had a slow and fragile start to today after the fun of the Current After Party last night... Bourbon Street is a lively place! 🍻

Dale Lane @dalelane.co.uk · 5d

I'm in New Orleans for #Current25

My presentation this morning was a high-level look at the four most common patterns for how people bring AI into event-driven architectures

From Event Streams to Smart Streams : Powering AI / ML with your Kafka topics

In this series of posts, I will outline the most common patterns for how artificial intelligence and machine learning are used in event driven architectures. I'm at a Kafka / Flink conference this we...

dalelane.co.uk

1

Dale Lane @dalelane.co.uk · 6d

used the flight to MSY to watch the new Superman film - really good - best superhero movie I've seen in a while

Dale Lane @dalelane.co.uk · 6d

Haven't made it to New Orleans yet... yesterday was a bit of a comedy of errors with mechanical and staffing issues meaning my flight took off *8 hours after boarding*, and left me stuck at JFK overnight.

Take two today. I just need to get from JFK to MSY... don't let me down again, Delta! 🙏

1 1 2

Dale Lane @dalelane.co.uk · 7d

I jinxed myself with this... we still haven't taken off yet, 7.5 hours after I boarded 😞

Dale Lane @dalelane.co.uk · 8d

on my way to New Orleans for Current - first stop: JFK
current.confluent.io/new-orleans

New Orleans

Current is the biggest event in data streaming. Over two days in New Orleans, developers, data practitioners, and tech executives will come together to shape the future of Apache Kafka, Apache Flink, and real-time data.

current.confluent.io

1 1

Dale Lane @dalelane.co.uk · 10d

Today, I played my 100th video game this year.
backloggd.com/u/dalelane/

That's a lot, right? I haven't tracked my gaming so closely before this year, but that feels like a lot.

dalelane's profile | Backloggd

Keep a virtual backlog of your video game collection, then rate and review the ones you've played to share with your friends!

backloggd.com

Dale Lane @dalelane.co.uk · 10d

no rain, so I've strapped my umbrella to the side of my bag

partly so I don't have to carry it

mostly so I can feel like I'm a ghostbuster

photo of the ridiculous combination.

it doesn't actually look much like a ghostbusters proton pack but i can dream

2

Dale Lane @dalelane.co.uk · 13d

I've written a new worksheet for Machine Learning for Kids - using #Scratch to give a simple interactive introduction to LLM benchmarks, and how people choose language models for AI projects

Introducing LLM benchmarks using Scratch

In this post, I want to share a recent worksheet I wrote for Machine Learning for Kids. It is perhaps a little on the technical side, but I think there is an interesting idea in here. The lesson b...

dalelane.co.uk

1

Dale Lane @dalelane.co.uk · 14d

watching my daughter run the Great South yesterday has got me looking for a run to sign up for... I'm tempted by the idea of running around a castle in the dark, mostly because it's so different to any run I've done before

www.maverick-race.com/products/the...

The Maverick Dark Castle Trail Running Event, Corfe Castle, Dorset

www.maverick-race.com

Dale Lane @dalelane.co.uk · 15d

curious indeed
www.amazon.co.uk/dp/1916051855

The Curious Case of Faith & Grace (The Trial Trilogy) : Lyons, David B.: Amazon.co.uk: Books

www.amazon.co.uk

Dale Lane @dalelane.co.uk · 16d

in Southsea for the Great South Run

(I'm not running - I'm here to cheer Faith on, in her first organised run)

Dale Lane @dalelane.co.uk · 17d

pleasingly (and partly why I chose this task for the worksheet) the performance doesn't solely correlate with the size and complexity of the model

screenshot from Machine Learning for Kids, showing how the language models I offer compare in size (both in MB and in number of parameters)