Mathew Thomas

Episode#6 - AI Investment Bubble - Is there one? This podcast delves into the current state of the AI market, exploring whether there is an AI bubble and the implications of recent investment trends.

open.spotify.com/episode/1FYm...

Most frustrating experience at Mumbai International airport with @lufthansa airlines. They keep postponing the flight departure and no announcements on what is causing the delay. Almost 15 staff roaming around lost and zero customer appreciation to explain why.

MiniMax M2 & Agent: Ingenious in Simplicity MiniMax M2 was released on Monday 27th October by MiniMax, a Chinese AI lab founded in December 2021. It's a very promising model. Their self-reported benchmark scores show it as …

MiniMax M2 is the new "most intelligent" open weights model (according to Artificial Analysis) - the MIT licensed weights are just 230GB and it appears comparable to Sonnet 4, while priced closer to Gemini 2.5 Flash. Notes here, including a new LLM plugin:
simonwillison.net/2025/Oct/29/...

I had a similar thought. At least companies will now attend to resiliency/ha/dr, especially on the home mattress overheating example.

The question is - what will be the change impact when we are on the other side of this. And don't bother trying to time this. Best you can do is embrace and retrain yourself for the other side of this

Just like the late 90s Internet bubble, the AI bubble is real and will burst. When that happens some companies will go down and retail investors will hold the losses. But AI is here to stay, just like the Internet when it was reborn after the dot com bust.

And God forbid there is a production bug in the LLM generated code...maybe call in another LLM at 3am?

As per some experts the future involves humans not touching the code. Assuming these tools get to that level of maturity to solve complex enterprise systems, get ready to rewrite these systems every year as the models get better.

Just another morning at Munich international airport!

🎯

25+ minutes of taxing for landing and takeoff at ORD Chicago airport is a test of human patience. Reminder to self to avoid this airport.

Claude Sonnet 4.5 is probably the “best coding model in the world” (at least for now) Anthropic released Claude Sonnet 4.5 today, with a very bold set of claims: Claude Sonnet 4.5 is the best coding model in the world. It’s the strongest model for building …

Wrote up my initial impressions of the brand new Claude Sonnet 4.5 - I think it may live up to Anthropic's claims of being the "best coding model in the world", for the next few weeks at least!
simonwillison.net/2025/Sep/29/...

Exploring multi-agent systems 🤖 with LangGraph and build a research assistant 📚 blogs.justenougharchitecture.com/building-mul...

So many people trade Mag 7 stocks rather than owning them, says Jim Cramer 'Mad Money' host Jim Cramer talks what is ahead for the Magnificent Seven.

When Carmer says to hold on to Mag 7 stocks then I want to do the opposite.

"So many people trade Mag 7 stocks rather than owning them, says Jim Cramer"

flip.it/kkwkmz

StubHub IPO access on Robinhood. Get in line if you want to be rich after being awarded 1 stock by RH !!

What is Mistral AI? Everything to know about the OpenAI competitor | TechCrunch Mistral AI, the French company behind AI assistant Le Chat and several foundational models, is officially regarded as one of France’s most promising tech

What is Mistral AI? Everything to know about the OpenAI competitor

flip.it/T6UPSi

Exploring the essential skills engineers need in an AI-driven world! Discover how adaptability, continuous learning, and embracing a problem-solving mindset are reshaping the engineering path.

#CareerGrowth #AI

spotifycreators-web.app.link/e/hpyv1zn9oWb

#unitedairlines please get rid of this design where the entertainment screen controls are on the armrest.

My thanks to @scalzi.com and @yoran.com for speaking to me about their experiences for this piece.

Generating images that have text content is still broken in GPT-5. So many words are misspelled and then some gibberish where one cannot even make out the text.

Is this new info? I use Claude to draft and update podcast scripts. It makes up fictitious quotes all the time. I have to tell it to stop. I will ask, and it will say yes, this is a fake quote, and also apologize. Ridiculous.

Episode#4 - Open Source AI: The Three-Way Battle for Tomorrow’s Intelligence Mat's Tech Thoughts! · Episode

Episode#4 - Open Source AI: The Three-Way Battle for Tomorrow’s Intelligence
open.spotify.com/episode/3gG6...

I am looking forward to the day where I can be this honest. There is so much to say!!

New podcast "When AI Goes Rogue". The recent #Replit database deletion incident serves as a perfect case study for the broader challenges facing our industry as #AI development #vibecoding tools rapidly evolve.

spotifycreators-web.app.link/e/zapqf8nKdVb

A language model built for the public good ETH Zurich and EPFL will release a large language model (LLM) developed on public infrastructure. Trained on the “Alps” supercomputer at the Swiss National Supercomputing Centre (CSCS), the new LLM ma...

ETH Zurich will be releasing an open-source language model designed for public benefit.

Hoping to get my hands on this to test soon.

ethz.ch/en/news-and-...

My second podcast (forgive the production gaps; it's me figuring out podcasting), this time on # AIAgent-based coding and vibe coding. #AIAgent #vibecoding

spotifycreators-web.app.link/e/geskbu9iFUb

Mat's Tech Thoughts! Podcast · mattazoid · Short podcasts on anything I find interesting on Tech. With over 20 years of experience, including a few Fortune 100 and startups, I hope I can share some of my knowledge with ot...

My first podcast ever. It took a while getting here, and the editing is not the best, but hey, this is episode #1, so I forgive myself. Topic is on continuous learning - open.spotify.com/show/65Ht8W5...

Building Java Applications with LangChain4j & Spring -
blogs.justenougharchitecture.com/building-jav...

As I delve deeper into AI Agent code assist tools, I cannot help but feel that a new kind of programming language will be developed for this AI code development world we have entered, as opposed to all the ones we use, which were designed for humans to write & understand. Thoughts?

Zero Trust and why we need to shift to that mindset and implement it sooner rather than later.
blogs.justenougharchitecture.com/zero-trust-s...