lukerm
banner
zl-labs.bsky.social
lukerm
@zl-labs.bsky.social
Innovation Tech Company for Machine Learning and AI
https://zl-labs.tech
https://parallellm.com
Moonshot AI just released Kimi K2 ... with Thinking. Now, it's available for use in Parallellm!

parallellm.com

This huge model has made a splash, achieving SOTA on several agentic benchmarks, such as Humanity's Last Exam and 𝜏²-Bench Telecom.

Another great one added 👍

#ai #llm #parallellm
November 7, 2025 at 3:04 PM
A few weeks ago, AWS went down. Hard.

Here's Fireship's funny post-mortem video of how it happened:

www.youtube.com/watch?v=HqL0...

A misconfigured DNS brings down half the internet 🤦‍♂️

#aws #fireship #fail
US-EAST-1 is humanity’s weakest link…
YouTube video by Fireship
www.youtube.com
November 5, 2025 at 3:33 PM
DeltaDB.

"A collaborative workspace where humans and AI agents work together across a range of time scales, with every insight preserved and linked to the code forever."

zed.dev/blog/sequoia...

This could be interesting...

#ai #git #zed
October 28, 2025 at 8:39 AM
A fascinating read about using Claude unabated on Simon Willison's weblog:

simonwillison.net/2025/Oct/22/...

The paradox of using it whilst staying safe.

#ai #claude #sandbox
Living dangerously with Claude
I gave a talk last night at Claude Code Anonymous in San Francisco, the unofficial meetup for coding agent enthusiasts. I decided to talk about a dichotomy I’ve been struggling …
simonwillison.net
October 23, 2025 at 5:46 PM
I made demo video to walk you through how to use Parallellm

www.youtube.com/watch?v=tUmH...

#ai #parallellm
Parallellm: Product Demo
YouTube video by Parallellm
www.youtube.com
October 17, 2025 at 2:44 PM
How to use n8n, with a Fireship twist. I've fast-forwarded to the demo.

www.youtube.com/watch?v=bS9R...

It's even open source!

#workflow #oss
n8n will change your life as a developer...
YouTube video by Fireship
www.youtube.com
October 17, 2025 at 10:17 AM
Tuesday brought lots of excitement as pi-Python (3.14) - or "Pithon", if you will - was released.

It's all about the colour: syntax highlighting straight in the REPL!

Check out the difference between versions 3.13 and 3.14!

#python #pithon
October 10, 2025 at 2:03 PM
Beta launch of Parallellm 🚀

parallellm.com

I'm absolutely thrilled to announce the launch of Parallellm in Beta - supercharge your AI conversations in parallel!

🫵 Beta testers needed! Try it out the platform for FREE, and share your feedback! 🙏
Parallellm
Large Language Models, Supercharged in Parallel
parallellm.com
October 2, 2025 at 1:33 PM
Sora 2 is out, and here's the masterplan:

"The majority of users ... should feel that their life is better for using Sora that it would have been if they hadn’t.

If that’s not the case, we will make significant changes..."

blog.samaltman.com/sora-2#:~:te...

Let's see how that plays out ...

#ai
Sora 2
We are launching a new app called Sora. This is a combination of a new model called Sora 2, and a new product that makes it easy to create, share, and view videos. This feels to many of us like...
blog.samaltman.com
October 1, 2025 at 7:46 PM
A great post-mortem analysis of Windsurf's demise:

ethanding.substack.com/p/windsurf-g...

"
Claude Code charges $5/day ($150/month) as a "good proxy for how much those 'should' cost," ... Cursor charges $20/month. Windsurf charges $15/month. They're selling dollar bills for quarters.
"

#ai
windsurf gets margin called
what the windsurf sale means for the ai coding ecosystem
ethanding.substack.com
September 19, 2025 at 3:58 PM
Then and now ... the state of the code in 2025. Times are changing!

(Comic from Andrew Ng's newsletter)

#ai #agentic
September 5, 2025 at 1:58 PM
Reposted by lukerm
Build a fully-local voice assistant via LangGraph + MCP servers with no subscriptions, no cloud, just fast, on-device smarts that actually work. The future of personal AI is private and modular.
#AI #VoiceAssistant #LangGraph #MCP #Privacy #ModularAI #LocalAI
Using LangGraph and MCP Servers to Create My Own Voice Assistant | Towards Data Science
Built over 14 days, all locally run, no API keys, cloud services, or subscription fees.
towardsdatascience.com
September 5, 2025 at 12:01 PM
A great talk by Andrej Karpathy on the "software update" in light of LLMs and friends.

We're currently in the 1960s era of computing when the computers were massive and centralized.

Well worth a watch!

www.youtube.com/watch?v=LCEm...

#ai #llm
Andrej Karpathy: Software Is Changing (Again)
YouTube video by Y Combinator
www.youtube.com
August 25, 2025 at 11:55 AM
I'm excited to announce ZL Labs has a new partnership with Lanboss AI ! Read more about it below 👇

zl-labs.tech/post/2025-08...

#ai
ZL Labs Partners with Lanboss AI
I am delighted to announce that ZL Labs is partnering with Lanboss AI, helping them on their mission to deliver secure and AI-compliant products in commercial and education spaces. As a part of this v...
zl-labs.tech
August 6, 2025 at 9:12 AM
AI-Powered XKCD Comic Finder! Read more about it on my blog 👇
bit.ly/zl-labs-xkcd...

Demo available here:
xkcd-finder.com

#AI #MachineLearning #SemanticSearch #RAG #xkcd
XKCD Finder
Retrieving XKCD Comics with Semantic Search and RAG
bit.ly
July 18, 2025 at 6:31 PM
A programmable desktop robot for about £200. It looks like fun! 🤖

techcrunch.com/2025/07/09/h...

#ai #huggingface #robot
Hugging Face opens up orders for its Reachy Mini desktop robots | TechCrunch
Hugging Face is releasing two versions of its desktop robot meant for designing and coding AI applications.
techcrunch.com
July 12, 2025 at 6:06 AM
This map of GitHub is awesome! I guess we're somewhere around the Learning Cortex 🧠

Link: anvaka.github.io/map-of-github

#ai #machinelearning #github
June 20, 2025 at 4:21 PM
An amazing, thoughtful and philosophical question about AI, asked by a young person most likely to be affected by the outcome 👏

"If you start talking to AI, if AI was in a human body, and talks and acts like a human, and you become friends with it, would it count as social interacting?"
June 11, 2025 at 5:42 AM
In 20% of cases, AlphaEvolve improved the previously best known solutions, making progress on the corresponding open problems.

😯

deepmind.google/discover/blo...

#ai
AlphaEvolve: A Gemini-powered coding agent for designing advanced algorithms
New AI agent evolves algorithms for math and practical applications in computing by combining the creativity of large language models with automated evaluators
deepmind.google
May 21, 2025 at 8:38 PM
SMBC on the future of coding 😂

www.smbc-comics.com/comic/use
Saturday Morning Breakfast Cereal - Use
Saturday Morning Breakfast Cereal - Use
www.smbc-comics.com
May 16, 2025 at 2:38 PM
OpenAI is changing its company structure to a Public Benefit Corporation. Jumped or pushed? Here's an interesting discussion on the history of this development.

www.youtube.com/watch?v=GwQV...

#generative #ai #openai
BREAKING: OpenAI ABANDONS Plan To Go For Profit
YouTube video by Matthew Berman
www.youtube.com
May 6, 2025 at 4:54 PM
The Parallellm Pump is a new developer power tool for comparing responses from LLMs in parallel. Read more about it on my blog here 👇

bit.ly/zl-labs-parallellm-pump

#genai #llm #ai #claude #chatgpt #deepseek #ml
Parallellm Pump
A power tool for comparing LLM responses
bit.ly
April 7, 2025 at 7:48 AM
[Sound on] I'm speaking with the V2V MoshiVis model in the video (NOT the purple stick figure 🙂)

MoshiVis holds much promise, but there are evidently a few kinks to be ironed out. Still, a great technology in the making!

#ai #ml #genai #transformers #moshi
April 3, 2025 at 4:28 PM
Anthropic's Claude is great at coding, but does make mistakes.

Part 2 (see image) suggests that you can use a lambda function to combine columns during a pandas aggregation.

This is confabulation! As demo'd in the image.

The rest is correct. Go carefully!

#GenAI #llm #ai #claude #pandas
March 25, 2025 at 5:33 AM
Does Claude 3.7 really "enable" bioweapons acquisition? An interesting RCT by Anthropic studies whether access to their LLM makes this possible - thankfully it doesn't! 😅 Participants made too many critical errors.
March 6, 2025 at 4:34 PM