25 articles on ai research.
AI Research
AI agents without memory are stateless. Every session starts from zero. For a production system where accumulated experience matters, that is not a limitation to accept — it is a problem to engineer around.
AI Research
A single AI agent attempting complex, multi-stage work will fail more often, and more opaquely, than a well-decomposed team with clear interfaces. The decomposition decision is not technical — it is architectural, and it determines everything downstream.
AI Research
Trust in an AI system is not declared. It is designed. The security architecture of a production multi-agent system — approval gates, scope limits, credential handling, boundary checks — is where trust is either built or not built.
AI Research
Routing every task to the most capable model is the most expensive approach and not always the best one. Task-appropriate model selection, based on output requirements and cost profile, is one of the more underrated disciplines in AI operations.
AI Research
Running 24 AI agents in production raises an immediate question: how do you know if any of them are doing good work? Our answer is a three-dimensional scoring system, an exemplar library, and a training loop that gets better with every failure.
AI Research
AI agent failures are not random. They cluster into six recognisable types, each with a distinct diagnostic signature and a different fix. Recognising which type you have is faster and more reliable than debugging from first principles every time.
AI Research
Most AI agent failures are not model failures. They are specification failures — the agent was not told precisely enough what to do, what not to do, and how to handle the boundary cases. Fixing this before deployment is considerably cheaper than fixing it after.
AI Research
I write for business readers, if you would like the math and science I have linked the papershereandhereand at the end of the article. Imagine Your Brain as...
AI Research
Building the Mechanical Parrot 1 | Is The Lid of Pandoras Box Wide-Open? In relation to both Quantum and AI, Pandora’s box is no longer ajar; it’s on the...
AI Research
Imagine attempting to squeeze every volume from the great library into Frodo’s tiny little backpack. A few books might fit, but soon the seams burst,...
AI Research
I think I may have been reading Quantum papers for too long, I am starting to fall deep down into a chain of rabbit holes and instead of a trippy smiling...
AI Research
Picture this: you're cycling through the hills of quantum technology, trying to power your way uphill, downhill, and around tight corners. Unlucky for you...
AI Research
I write for a business audience, if you are a Physics boffin I suggest you tuck straight into this paper here. For the rest of us.... A team at Google...
AI Research
For business leaders, innovators, and strategists: if you want a glimpse into what could reshape industries before 2030, this is worth your next 5 minutes....
AI Research
I write for a business audience, if you are a PHD you may just want to hit the paper here, clearly it has nothing to do with Peanut Butter, ish...PB=MBL For...
AI Research
Scientific boffin - skip my narrative and read the paper here. For the rest of us.... In yet another month of grand claims from quantum companies, it seems...
AI Research
Neural Networks and Grassmann Algebra: Strap in. I write for business people, some may think that Grassmann Algebra is a new type of craft ale. If you are a...
AI Research
Scientific Code Boffin? - Read the paper here, I write for non technical business people, this article may annoy you. For everyone else that needs to...
AI Research
Read the scientific paper here, or take it easy with a no math explanation. Fast Amplitude Embedding for Quantum Machine Learning So let's imagine that you...
AI Research
I have deliberately avoided jumping on the announcement bandwagon over the past weeks. Especially regarding this one (Read the paper) to discuss D-Wave For...
AI Research
If you are a Quantum Scientist then click to these papers. [A Review of Barren Plateaus in Variational Quantum Computing] [Large Language Models Can Help...
AI Research
[ I write for a non technical business audience, but physics boffins can check out the full research paper here:[arXiv:2502.21055v1 [quant-ph]...
AI Research
Ever Heard of the "Too Much Talking" Problem? Imagine you ask a friend for directions to the nearest shop, and instead of a quick "Turn left, then right,"...
AI Research
Imagine trying to fit an elephant into a suitcase. Tough, right? That’s kind of what AI researchers are doing—making massive AI models smaller and more...
AI Research
NB: If you are a highly technical PhD type software person, best to skip this and go straight to Dr Javier Mancilla Montero, PhDpost hereand then readthis...