Artificial Intelligence
Having worked on developing Machine Learning skill definitions and L&D pathway recently, in this blog post I have tried to explore the evolving role of test engineers in the era of machine learning, highlighting the key challenges ML brings, strategies for effective model evaluation and finally tried to outline a roadmap for developing the necessary skills to excel in ML model testing.
Artificial Intelligence
Our team set out to investigate if we could use Binoculars, a method of detecting LLM-generated text, to detect AI-written code, and learnt some valuable lessons on the importance of data quality along the way.
Artificial Intelligence
As the government progresses from prototype to production to ongoing operation with AI-assisted services for UK citizens, how can it minimise the risk of replicating structural biases? In this blog post, I’ll explore key elements of what’s involved in ensuring that services are as representative, fair and impartial as possible.
Artificial Intelligence
I reflect on recent discussions with civil servants and our own research to consider how AI might increase productivity and offer new capabilities. At the same time, I’ll explore the necessary checks and balances on how far AI should be applied in delivering services to UK citizens.
Artificial Intelligence
Data is important to every element of the safe and successful harnessing of AI. For the government, data challenges are not new, nor are they insurmountable. In this post, I’ll look at some of those challenges across the data lifecycle, and share insights from our recent work and conversations with government about practical ways forward.
Artificial Intelligence
Government departments are not alone in struggling to move artificial intelligence (AI) prototypes into production. Here, I reflect on some of the reasons for this and share some insights from our client work which I think point towards how the government can harness AI with greater confidence.
Artificial Intelligence
Generative AI has a Sustainability problem – across environment, cost and continuing to meet expectations. Many companies are racing to implement GenAI in their projects, lured by its hyped potential to revolutionise industries. However, in applying GenAI to enterprise implementations, I am seeing first-hand the sustainability challenges threatening to implode the first generation of this technology. This blog talks about what I hope will rise from the ashes of such an implosion.
Artificial Intelligence
Many of us are now using Generative AI to produce code, but what impact is this having on code quality? By analysing the data contained in PRs, our team has built a tool to investigate this.
Artificial Intelligence
As we keep on building the InferGPT chatbot, the task of crafting prompts is one that our team is experimenting with. In this blog I'll explore some of the challenges of prompt engineering from a Linguistics angle.
Artificial Intelligence
In this blog, I describe creating InferGPT, a personalized chatbot, using Python, FastAPI, and React without the use of LangChain. Our backend handles complex queries with a multi-agent system and Neo4j database, aiming to provide tailored responses.
Artificial Intelligence
This blog series reports on our R&D project exploring the potential of LLMs to enhance accessibility testing.
Artificial Intelligence
This blog demonstrates how to generate graph database queries using a large language model. This allows users to ask natural language questions in order to understand their data presented in a knowledge graph better! This can be done without extensive understanding of the querying language.
Artificial Intelligence
Exploring LangGraph and its applications in the context of multi-agent Large Language Model (LLM) applications.
Podcast
In this episode, I'm joined by Jess McEvoy and Peter Chamberlin for a discussion about the UK government’s approach to addressing AI challenges with its pro-innovation mantra, and whether this creates the right environment for success.
Artificial Intelligence
Using an example project this article explores the possibilities of integrating a Large Language Model with custom logic to improve its utility and power.
Artificial Intelligence
Can AI write software? My team at Scott Logic wanted to find out whether LLMs could generate JavaScript code to draw a D3 chart in a web page. Read about how it went in this blog post.
Artificial Intelligence
.JPG)
Reflections from the recent 2024 Public Sector Innovation Conference in London, on the topic of Artificial Intelligence
Artificial Intelligence
Accurate weather forecasting helps us make decisions every day, from the trivial to the critical. I take a look at the technology that goes into predicting the weather, and consider how the AI revolution could change that.
Artificial Intelligence
When we couldn't find a tool to meet our prompt engineering needs, we decided to create a universal tool that allows anyone to build their own – all within the familiar spreadsheet environment. An Excel add-in that lets you use OpenAI models in Excel formulas.
Artificial Intelligence
In this blog, we will be sharing our feedback on the free fast.ai course - that has been praised by many for its hands-on and top-down approach - along with a list of additional material that we found helpful in fully understanding the course.
Artificial Intelligence
Recently I’ve been learning about Neural Networks and how they work. In this blog post I write a simple introduction in to some of the core concepts of a basic layered neural network.
Artificial Intelligence
Developing LLM applications using Test-Driven Development principles can be expensive due to the incurred cost of every usage. In this post, we discussed how our attempt to tackle the inherit non-deterministic nature of LLMs increased our operational costs and the techniques we employed to minimise them.
Artificial Intelligence
We’ve undertaken experiments to explore the impact of GenAI tools on developer productivity, revealing a 37% improvement in productivity (speed), however, this result is a misrepresentation of what it means to be productive as a developer. This article delves more deeply, beyond punchy metrics, to explore the overall experience of working with these entirely new tools. We discuss where these tools are most effective, the challenge of quality code, the learning curve and much more.
Artificial Intelligence
In this post, we walk through the steps required to create your own DiffEdit pipeline, allowing you to use generative AI to edit parts of existing images.
Artificial Intelligence
An introduction to the random forest machine learning model, an older, yet interesting approach based on decision trees.
Artificial Intelligence
In this post, I provide an overview of the capability models that are out there, summarise the distinct capabilities required for AI/ML projects, and make the case for what I believe to be the most important capability your organisation will require.
Artificial Intelligence
In the light of OpenAI’s system outage, a ripple of unprepared AI powered systems crashed across the world. How can we best secure our LLM based applications against such occurrences?
Artificial Intelligence
I had the pleasure of attending Cvent Connect, a prominent conference for events and marketing professionals held in London last week. Can you guess what the hottest topic for debate was? You're right, it was none other than AI and Tech!
This time it was in the form of the debate of People vs Tech in copywriting and marketing. Read this blog to find out which/who emerged victorious.
Artificial Intelligence
Have you ever tried explaining something to a child, who pretends to understand, but in reality doesn’t have the foggiest what you’re saying? That’s the experience I’ve had with LangChain agents.
Artificial Intelligence
Our journey of learning about Prompt Injection Attacks, securing Generative AI, and how we became undercover spies on the way.
Artificial Intelligence
While the latest AI technologies may be new, your best next steps in harnessing them are tried and tested, and you already have many of the skills you need. I set out in this blog your best next steps to get ready for AI in Government.
Tech
Understand how Google and other search engines use LLMs to gain insights into the semantic meaning of the language in search queries using embedding and cosine similarity.
Artificial Intelligence
I was part of a Scott Logic team that developed and delivered a Proof of Concept (PoC) for a major retail bank which demonstrated high potential business value and showed how GenAI could be reined in as part of a range of options to mitigate risk. In this post, I provide an overview of how it worked.
Artificial Intelligence
ChatGPT is put through its paces to see how the AI fares as a productivity tool for developing a small project. We look at where it helps, where it doesn't, and where AI tooling might go in the future.
Artificial Intelligence
AI tools are being trialled in many areas including software development, with huge potential to transform ways of working. What might the impact of AI be on software architecture? In this post, I make some predictions and argue that software architecture may become more data-driven, with prototype architectures and architectural diagrams being generated from the actual code itself.
Artificial Intelligence
Generative AI is moving at an incredible pace, bringing with it a whole new raft of terminology. With articles packed full of terms like prompt injection, embeddings and funky acronyms like LoRA, it can be a little hard to keep pace. For a while now I've been keeping a notebook where I record brief definitions of these new terms as I encounter them. I find it such a useful reference, I thought I'd share it in this blog post.
Artificial Intelligence
This post offers a proposed architecture for mitigating the risks of deploying Generative AI technologies in complex enterprises, particularly when used in highly regulated environments such as Financial Services. As well as the proposed architecture, I provide a primer on the business risks of Enterprise GenAI along with the technology risks and challenges.
Artificial Intelligence
LangChain has become a tremendously popular toolkit for building a wide range of LLM-powered applications, including chat, Q&A and document search. In this blogpost I re-implement some of the novel LangChain functionality as a learning exercise, looking at the low-level prompts it uses to create these higher level capabilities.
Artificial Intelligence
It’s been over a year since I last blogged about OpenAI. Whilst DALL-E 2, ChatGPT and GPT4 have grabbed all of the headlines, there were a lot of other interesting things showing up on their blog in the background. This post runs through just over six months of progress from Sept 2021 - March 2022.
Artificial Intelligence
The buzz and excitement around generative AI is continuing to grow as their capabilities rapidly expand. However, their ability to generate content is just the starting point. From my perspective, the emergent reasoning capability, coupled with their intelligent use of tools, is what will make this technology truly transformational, and will mark the start of a new technology epoch.
Artificial Intelligence
It is quite common for blogs, or news / content distribution websites, to organise their content around categories of tags. However, this approach is time-consuming, and from measuring behaviours, it doesn't seem to do much to encourage browsing. In this blog post I use the new OpenAI embedding API to create highly targeted article recommendations, resulting in a four-fold (x4) increase in engagement.
Artificial Intelligence
·
Video
In this talk, I share the key things AI can learn from Neuroscience, and why these two disciplines need each other in order to progress successfully.
Artificial Intelligence
You can tell powerful stories with data, but so often we are faced with raw data (albeit beautifully presented), and are left to create our own narratives. My previous attempts at turning data into stories have been time consuming and underwhelming. In this post I demonstrate how GPT-3, a new and advanced language model, can construct engaging and unique stories from user-specific data, with relative ease.
Artificial Intelligence
·
Video
I talk about GPT3, one of the massive and very impressive AI models to come out of OpenAI. I quickly cover off the basics of how it works and where its creativity comes from, before accidentally delving into existentialism.
Artificial Intelligence
The OpenAI Codex JavaScript Sandbox and the API which backs it, have (rightly!) generated a lot of interest and comment in the development community. When faced with something so different to what we expect, it's easy to assume it's all magic. In this post I'll peel back some of the layers and try to describe what's going on.
Artificial Intelligence
...or what I should have known before I jumped in and started playing around. In this post I cover few shot learning and fine-tuning.
Artificial Intelligence
...or what I should have known before I jumped in and started playing around. In this post I cover the basics of tokens, the model and prompt design.
Artificial Intelligence
·
Video
Last month GitHub released Copilot, an AI powered tool that provides surprising accurate suggestions ranging from a few lines of code to entire functions. Copilot uses a vast and powerful AI model that is trained on billions of lines of open source code from GitHub. This, combined with the context provided by your code and comments, allows it to provide amazingly accurate suggestions. This is no simple autocomplete, this is pair programming with a robot!
Artificial Intelligence
I've been lucky enough to be granted access to the preview of Copilot, GitHub's new 'AI pair programmer' that is causing quite a stir. This blog post shares my initial thoughts and experiences with this tool. In brief, I'm stunned by its capabilities. It has genuinely made me say 'wow' out loud a few times in the past few hours, not something you expect from your developer tooling.
Artificial Intelligence
Recently I've been doing a bit of research on machine learning and particularly TensorFlow and Keras. This blog provides a brief introduction to the tech, and then a simple tutorial using both TensorFlow and Keras, exploring the differences between the two.