Matthew Griffiths

Well, of course it constitutes a merger in an attempt to go around antitrust and anti-competition laws. Big tech is already providing the computing infrastructure and most of the funding to LLMs, so what's left is what products the teams were working on, the computers they were working with, talent and experience. Why buy the company when you can poach the talent? Especially when probability is high that an acquisition would be declined by regulators due to anticompetitive impact on markets. https://lnkd.in/g_6q45FE

5

In London for London Tech Week, and working on getting AI into production? Our next London NLP Meetup group evening event on Wednesday June 12th has amazing speakers from Weights & Biases, LlamaIndex and Weaviate. Sign up for free here, details below: https://lnkd.in/e5PWG7au Solving the challenges of using LLMs in production with financial services data If you are building NLP pipelines for processing financial services data, you will know how hard it can be to manage vector databases in production, reliably process large tables, and evaluate your LLM performance for reliability and accuracy. At this event during London Tech Week, we’re delighted to have three amazing experts share their cutting edge expertise: Topic: Solving the challenges of scaling vector databases in production. Speaker: JP Hwang from Weaviate His talk and demo will cover advanced topics such as quantization, model selection and database sharding. Topic: Solving the challenges of handling doing RAG on top of financial data. Speaker: Pierre-Loic Doulcet from LlamaIndex His talk and demo will cover new techniques for better handling of financial data with a focus on tabular data and charts. Topic: Solving the challenges of experimentation and evaluation of LLMs for production. Speaker: Ryan McConville from Weights & Biases His talk and demo will cover approaches for LLM experimentation and evaluation. By the end of the meetup we will have gained practical, actionable insights and takeaways that can accelerate development and deployment of NLP in production.

24

2 Comments

I have been enjoying the improvements in Claude Ai.. Here are my thoughts. https://lnkd.in/evUT_psn

1

“We’re still not even at cat intelligence yet, as a general system,” - On 9 July, at the Future of Britain Conference 2024 – hosted by the Tony Blair, here’s what Demis Hassabis, CEO of Google DeepMind said while answering a question about DeepMind’s progress in artificial general intelligence (AGI). A must-watch perspective of Demis Hassabis with Tony Blair for all AI investors, customers, practitioners, podcasters and users: https://lnkd.in/dV4Sx5sg If Google with all the access to 100s of billions and 1000s of PhDs (and most of the data in the world) says this is where they are currently when it comes to AGI, the real question is: Is there a viable path to AGI? Check out INSA (Integrated Neuro-Symbolic Architecture) from AGI Pioneer Peter Voss: https://lnkd.in/dVbQJNHC Key question: Would you continue to invest more into #GenAI or consider investing into alternative approaches like #INSA (Integrated Neuro-Symbolic Architecture)? Here is Peter Voss chapter on the direct path to AGI from 2002, extracted from the Artificial General Intelligence book: https://lnkd.in/gxUNriaa https://lnkd.in/dsKWzJPB

12

6 Comments

Deciphering AI's Inner Workings: Scaling Sparse Autoencoders to Claude 3 Sonnet - A good read! https://lnkd.in/eYuwg7_U

9

AI & Partners - How does the UK intersect with EU in terms of #AI regulation? This ground-breaking report from OpenUK helps to shed light on a question of growing significance. With EU #AI Act implementation imminent and questions of its extra-territorial nature front and centre, business leaders can benefit from these expert insights. Check out the post below for details. #AI

7

If AI Agents will be priced and monetized the way pay for professional services (hourly fees, retainer, project scopes), we'll finally get to see the full 180° 'Software-as-a-Service' turning 'Service-as-a-Software'. And why shouldn't they? Consumers buying preferences has a fashion to it - the pendulum swing, and we tend to go back to patterns we're familiar with. When AI Agents be fully capable of replacing certain human jobs, it only make sense we pay them the way we'd pay a human getting the job done otherwise (dollar value may be diluted, but foundation persist). Think AI agent doing design work for your business. $250/hr for similar output, 1000x faster. Are we at the verge of a new way to think SaaS valuations? What's the the biggest public professional services companies you know? With a little manipulation to their cost of doing business (assuming building and operating AI agents has different economics of scale than employing hundreds of real humans for the same job), we may get an early glimpse into our future.

36

6 Comments

A personal plea to the new Government to grab the opportunities for UK growth presented by AI. Before the election, Labour promised changes in AI regulation and development in the UK. Plans floated included partnerships with industry and academic institutions to harness AI’s potential, regulating the makers of the largest AI models, and a really interesting one to create a Regulatory Innovation Office to streamline regulatory decisions, which could be turbocharged with the help of AI. The last Government got bogged down in “AI Safety”, focusing on existential threats. Using UK know-how, and with the right backing and investment, we can make AI greener and more sustainable. This should form part of the proposed National Wealth Fund. The Government should also encourage AI adoption across the public and private sectors. What we should not do is allow Government to take all of its AI advice from big US tech players, nor should we allow public sector adoption of AI to be driven by the path of least resistance, defaulting to the same old tech stack.   It is great that companies like Microsoft, Facebook and Google invest in the UK, bringing jobs and resources, but if we do not pursue sovereign AI solutions in parallel, we will stay the “51st State”. Outside of the US, a lot of the big news in AI seems to be coming from Paris right now... Let's see detailed commitments to R&D funding, and concrete proposals for supporting UK talent. We can ensure that the UK leads in AI innovation and creates a fair and sustainable future. Let's work towards making the UK a global AI leader. #Sustainablity #AI #Innovation #PrimeMinister #Labour #UK 

10

2 Comments

https://invoked.uk/ - Embracing the change with our new homepage!! Integration of LLMs, conversational search and Azure based OpenAI models on any web or mobile platform. #AI #machinelearning #predictivemodelling #datascience #largelanguagemodels #openai At Invoked, while adapting to new technology, we ensure our offerings are technically sound.

4

1 Comment

The UK’s AI sector has been valued at $92bn (£72.3bn), larger than any other country in Europe, according to new data. The ‘UK Tech in the Age of AI’ report from Tech Nation found that in the first quarter of 2024, 16% of all UK VC investment went to AI start-ups. According to the report, the UK is home to more than 1,800 VC-backed AI start-ups. The data put the UK at the top of Europe for AI investments and fourth globally behind the US, China and Israel.

7

This is another useful proof point of the need for #AI re-skilling across all companies and organisations. Siobhan's list of her own re-skilling and educational activities is a tribute to her inquisitive mind and resilience. It's clear that there's some great material out there from the big tech players and leading universities. What is less clear is what you need to do - at a personal or organisational level - to re-skill and reposition yourself. There's simply so much out there. With this in mind, my co-founder Shivika Pratap and I are completely reshaping https://HeyDeesha.ai to focus on guiding careers in the context of the AI economy. We are doing our reshaping based on a taxonomy of roles that we see emerging. Please DM me if you are interested in what the taxonomy looks like. It's not just tech roles, BTW. We want to extend the taxonomy to Marketing, Sales, Support... and, one day, really deep specialist skills such as Medicine. We are also looking at how the pace of change in the workplace is affecting maternity leavers/returners. Kate Apaflo Awuku-Darko, Emmanuel Ihenacho

10

3 Comments

Bidtune AI wants to revolutionise the way businesses approach the bidding/tendering process. We are looking for advisors who are open to equity-based arrangements. We need capabilities in the following areas: - Fundraising strategy - Sales & Marketing - Artificial Intelligence/Machine Learning - Data strategy - Legal (especially on IP, data & AI) - Extensive bid management experience in various sectors - Connections with B2B Businesses (warm intros) If you are interested, drop me a message. 😀 #fundraising #sales #data #AI #MachineLearning #ML #data #GDPR #legal #startup #advisors #intros #IP #datastrategy #tech #phd #bidmanagement

22

3 Comments

🚨big news from Emperia 🚨 Virtual worlds are notoriously hard to build - especially when it comes to running a high-fidelity experience on the web. They require a lot of asset optimisation, compromising on quality. Custom coding for interactions and commercial functionalities. And large teams involved in deploying each experience, which often takes months to create and deliver. Over the last 5 years, we’ve solved all those problems for our own 3D teams. We’ve built the underlying technology that enabled Emperia design, build and publish virtual experiences at scale. We always focused on visual quality and performance, and built the platform with the end user in mind. Today we’re officially opening our tools up for the creative community to build on Emperia platform. Built on UE5, Creator Tools enable: ✨Render and publish cross-device virtual experiences straight from Unreal Engine with no asset optimisation required ✨Plug-and-play UI, enabling any 3D artist to create a fully functional experience with no dev support required ✨ Procedural generation tools and lighting presets to create stunning environments in minutes ✨Commerce modules to support e-comm integration for sales & marketing experiences ✨Customizable interactions from gamification to token-gated rooms with an SDK to expand functionality even further …and so much more! Shoutout to my Emperia team for making this happen, and our partners at Sawhorse Productions, Cominted Labs, Polycount.io, virtuall and many others who helped us refine our creator platform in beta over the past few months 🙏🏻 Simonas Holcmann Emma Preti Charlie Adams Jadir Lino Tomas Tamulenas Tomas Gavelis Zunaib Imtiaz Stephen Cherla Shahzaib Asim Marcelina Slabuszewska #UE5 #EmperiaCreatorTools #builtwithEmperia #3D #unrealengine #AWE2024

199

18 Comments

Can the UK public sector use AI? Yes, but pubic bodies have to follow the Generative AI Framework for HMG: https://lnkd.in/eDXvWrKf The issue is this is a very long document with a huge number of links out to other documents. On the one hand one might argue that effectively all it says is – follow hmg policy for implementing IT. On the other hand – there’s really nowhere else that this diverse and overlapping set of policies have been set out in one place. In the next videos I’ll dive into the Principles, Ethical themes and how from a practical perspective the document raises more questions than it answers. #ai #ethics #publicsectorai

7

Google DeepMind launched Gemma 2, an open large language model available in 9 billion (9B) and 27 billion (27B) parameter versions. It outperforms larger models, providing cost-effective deployment options. Core Innovations Gemma 2 features sliding window attention, soft-capping, and knowledge distillation. Integration and Compatibility Gemma 2 integrates seamlessly with major AI frameworks, supporting Hugging Face Transformers, JAX, PyTorch, and TensorFlow via Keras 3.0. It runs efficiently on various hardware, from gaming laptops to cloud setups. Deployment and Access Developers can access Gemma 2's model weights from Kaggle and Hugging Face. Starting next month, deployment on Vertex AI will be available, with model integration options in Google AI Studio and local environments using Gemma.cpp. Safety and Evaluation Google DeepMind implemented rigorous safety measures, including data filtering and comprehensive testing, to mitigate biases and risks in Gemma 2. Academic Support The Gemma 2 Academic Research Program offers Google Cloud credits for research use, with applications open through August 9.

43

3 Comments

The next generation of personalised digital services will use data ecosystems and AI to build fairer and more profitable services. This short article on Medium outlines a few thoughts about what they will look like.

7

I'm going to share the day I stopped trusting GoogleDrive and it wasn't that recent. Summer of 2021, I was working on a pitch of a superhero property for a game with a 'Hollywood' client. And I stopped trusting Google to hold my work docs when I noticed the level of surveillance advertising going on between Google and Amazon. Now I like and use Google and Amazon, but I expect work to stay with work. Don't sell me when I'm working or with what I'm working on. We use Microsoft Teams internally. And I have used and have, every communication, task tracking, and conferencing application app under the sun. So our client wanted us to move the pitch deck (PowerPoint) to Google Drive and eventually to Google Slides. I took the deck that we'd been sharing back and forth internally on MS Teams and that resided in Teams/Sharepoint and out it up onto my Google Drive. We then started a process of bringing it over from PowerPoint to Google Slides. This meant that the deck and its contents had touched Gmail, Google Drives and Google Slides. Within 24 hours of doing so I noticed that in my Google Chrome browser I was being served Amazon ads for purchasing copies of the comic this super hero property was based on. That's right. Google had mined some combination of Google Drive or Google Slides or the fact that sharing from them is done via Gmail notifications. One way or another I was staring at ads in my Google Chrome browser of the top secret superhero IP that we were working on. This was a pitch that due to having taken it out of Microsoft Teams, Sharepoint, and PowerPoint, it had been exposed to Google Drive and Google Slides. Also keep in mind that the Gmail share notifications used a filename with a codename but had a thumbnail from the cover of the deck showing the superhero. What was the point of entry for Google Ads and Amazon Ads? I don't know exactly One way or another, that fed into both Google and Amazon both serving me ads across the next 2 weeks while we finished that deck up with the client. And I never put another deck into Google Docs. And that was in Summer of 2021 long before this current push to bring everything into AI training. Now to be fair I was paying for MS Teams/Office 365. And I wasn't paying for Google Workspace, though the client was whose Google Drive it went into. And yes, this was all in the days of just regular old surveillance advertising for ads. I have zero desire to see AI training on our corporate strategy. And I refuse to put docs into a place where they'll be fed into Ads then or 'gen' AI now. For any corporate partner. Google, Amazon, Microsoft, Salesforce, Zoom, Adobe, etc keep your shit secure, or scare off corporate partners large and small, who don't want their ideas and work used either for Ads or for AI training. No corporation is too big or too necessary to stop users from migrating elsewhere. #userdata #dataprivacy #security #noads

79

25 Comments

In view of the recent news, I predict a surge in discussions around AI regulations by the imminent elected UK political administration. This has been at the epicenter of debates in the tech industry, with heightened pressure for the incoming government to keep pace with the escalating mutation in technology and the AI narrative. Upholding the spirit of insightful conversations, let's delve a little deeper: • Driven by a wave of technological evolution, UK's competitiveness in the global ecosystem largely depends on how well the state fosters innovation whilst successfully curbing the fruits of Artificial Intelligence. There's a need for finely balanced legislation framing AI uses and applications, creating an economic leverage that propels growth. • This symbiotic relationship between AI and economic progress solidifying our standing, engenders an unavoidable reality. The incoming UK government needs to pay substantial attention to this technology sector that's infiltrating every industrial and consumer domain, hence avoiding the danger of sliding down the competitive ladder. • In the words of tech leaders like Rupal Karia, the need for tailoring AI rules to meet the unique demands of businesses and industries cannot be overemphasized. In a world that now revolves around AI-related designs, stratagems must be woven with a tapestry of awareness, comprehension, and utmost moral values. My ultimate prediction, encompassing this discourse, is the incoming UK government will step up to the call, harness and prioritise AI technological advancement, charting an expansive growth trajectory for the UK, amid an increasingly digitalized planetary economy. The lessons to hold onto here? Mastery of technological innovation becomes futile without a collaboration of well-designed policies. We stand on the precipice of a digital era where embracing and catalyzing AI technology becomes essential to stay ahead in the race. Placing in perspective, optimists like myself ambitiously yearn for a vibrant future where industry leaders and political fronts are harmonically synchronized, advocating AI, and accelerating its legitimate proliferation throughout the sectors. #AI #ArtificialIntelligence #TechNews #UKpolitics

2