Key Insights from Google Cloud Next/IO & AWS Summit 2024: TLDR; It's all about AI & Agents.

If you're reading this on your commute, lunch, or listening on audio, buckle up. This deep dive explores the whirlwind of AI and ML advancements showcased at Google Cloud Next, AWS Summits, and Google I/O. We'll unpack the opportunities, challenges, and implications for businesses and consultants alike.

The recent Google Cloud Next 2024, then AWS Summits 2024 and Google I/O Developer Summit alike provided updates on the unbelievable Acceleration of Technology and opportunities these presented for the Innovation Landscape, with such explosive growth of Machine Learning model advancements.

This resulted [and continues to evolve] in a myriad of Generative-AI-using [#GenAI] Capabilities, as well as specialised Artificial Intelligence (AI) solutions.

JP PoV: This is a hot new area of trend vs hype and quite real innovation unlike any other. It touches on many business pain points and creates yet more opportunities for others. No Single Business or Industry Vertical - where cloud or technology operates - will escape AI's far-reaching grasp, and indeed consequences.

We are fully committing and crossing the point of what many call the AI evolution today, - these technologies are reshaping industries, redefining modern workflows, and setting new benchmarks in innovation.

It's all About AI

Google's Gemini AI, with its expanding context window to 2 million tokens, is poised to revolutionise how enterprises analyse massive datasets, along with it's matured MLOps Platform Vertex AI. That said, AWS is also leading the charge with advancements in SageMaker to democratise the ML Process, making it accessible to businesses of all sizes.

It's quite noteworthy that the general user adoption is growing by orders of magnitude. It's either voluntarily to cultivate the personal or professional Moat, or brought in kicking and screaming into AI-augmented services.

Think Google Workspace, and Microsoft Copilot, or interaction with Google Android Gemini AI Overlay or Bing that is powered by GPT4 - AI everywhere

Preface: Consulting for, and with, AI

The breakneck pace presents a unique challenge for consultancies and enterprises alike, with relevant ML/AI skills being a real shortage. The demand for expertise in AI and ML is skyrocketing with a million roles by 2027 in ML/AI.

There is a plethora of confusion, unhelped by, and yet a genuinely rapid, pace of innovation. Relevant: How-To Study Guide & Strategy to get yourself Machine Learning Certified with Google Cloud

JP PoV: This makes it challenging to keep up with the "art of the possible today" that quickly superseds itself a "digital commodity by tomorrow" both for Business organisations as well as consultancies, working on these solutions. This is the effect of "Generative AI is eating the Sofware that is Eating the world"

Optimising the solution for the problem today, which may not be a problem tomorrow, leaves many businesses and organisations seemingly dragging their feet in adopting Generative AI, unrealising the full potential in many MVP engineering projects.

JP PoV: To put it bluntly, - If you're an information business, and yet not actively exploring how to leverage GenAI as your competitive moat - You're putting yourself at Risk. If you are not exploring this. then Your Competitors would.

Welcome to my latest outlook, a follow-up in this ongoing transformational change and Impact across industries.

• Embracing Generative AI: A Technology Consultant's Perspective on the Exponential Age (March 2023)
• The need for frequent upskill in the age of GenAI (Sept 2023)

About The Big Event of the Year thusfar

I explore the biggest AI-related updates of H1 thus far, and I will give you my closing thoughts on where I see this AI wind is blowing.

The #GoogleCloudNext took place in April 2024 in Las Vegas, and was probably the biggest one yet, with over 35,000 in attendance. This new conference dwarfed the last GCNext that took place merely 8 months prior in San Francisco with circa 16,000 visitors. Back in 2023, this thing* ...Generative AI excitement was picking up in earnest with many promises made about the "future" in 2024, and all Cloud Vendors were selling GenAI as proverbial cotton candy.

This one had a different feel. All those 2023 GenAI promises are finally and steadily getting delivered, - the AI/ML Tech is less mysterious and empowering with a material backdrop to "Everything Cloud is powered by Gemini AI" was very much notable. PS Did I mention the conference was very much so about AI? ...yeh? You get the point.

This is an overarching theme across the whole tech sector. There is no escape.

My views are informed by the Google Cloud Next Conference, and AWS Summit that have taken place in London, alongside my own hands-on into ML/AI #GenAI use cases.

5 Key Takeaways

The rise of specialised AI agents signals a shift towards more collaborative, human-like AI interactions. This could revolutionise customer service, employee productivity, and even how we interact with everyday devices.

1. It's all about the Agents. (no, really AI stole the Cloud Show) The Generative AI ecosystem is evolving (IMHO) to focus more on dedicated, smaller models which are experts in specific tasks and activities, who are then supervised by other LLMs who are more general in nature to do the Big Picture Reasoning. Think of it like an expert of experts or Neo-Industrial Management structure but for the GenAI Agents.

Ditto, I am purposely mixing Retail and Enterprise relevant updates on this topic because it would imply, there will be a convergence moment where Users will interact with Agents more readily than "browse to a website" to access an enterprise service. Experiment, and Plan accordingly.

2. Enhanced ML Models and Tools: Both Google and AWS [an AWS Summit, London] unveiled new enhancements in their ML offerings, focusing on accessibility, scalability, and efficiency. Google's introduction of Vertex AI matched AWS's improvements to SageMaker, both designed to streamline the development and deployment of ML models.

TLDR; MLOPS is taking centre stage in this emerging Large Language Model (LLM) focussed ecosystem. Gemini is the new branding and main model name going forward (no Bard, no Duet-AI business) And Gemini v1.5 is accessible for enterprise use, an industry-leading context window.

But Wait - There is more. Gemini context window is becoming a whopping 2 Million tokens, later this year in 2024. This opens vast creative uses where enterprises may wish to take the entire repository in context, for analysis, in one go for example.

2. Rise of Personal AI: Generative AI was a hot topic, with Google's Gemini Flash and Project Astra (for Edge/Mobile use) as an Assistant model and AWS's Amazon Q, and deep learning competencies showing how businesses can generate dynamic media, create content, but also starting to offer ideas and views about user-personalised consumption of such capable modes.

From Chatbots (RAGs) to Personal AI Assistants - and just like that, the stage is set for the next evolution step change in Human - Device behaviour change. We go from HID(Human Interface Device) to HIAI(Human Interface AI).

3. Focus on Ethical AI: Amidst the technical showcases, there was a strong emphasis on ethical AI development. This reflects a growing awareness of the need for responsible AI that is equitable and fair. This big scope of effort where we need strong leaders to see this through, and where we need strong governance and a system of checks and balances to ensure resulting LLMs are safe, secure and regularly assessed, especially when serving the public.

Sidenote: The UK government has a Responsible AI Toolkit and a National AI Strategy, This crosses over into cloud vendors, and LLM-consuming and service-providing organisations alike.

JP PoV: We're on the precipice of really capable models, whether you follow the AGI narrative and risks, or not, and it's most imperative that Responsible AI work continues with yet more focus for the foreseeable future, as many emergent AIs are more capable than humans. You want your AI to be aligned with YOU and perhaps be more rational and ethical as a standard.

4. Developer Experience Supercharged: Both Google and AWS are making AI development more accessible than ever before. Google's Gemini is now embedded into every key topic of the cloud ecosystem, not just the Google Search, and Google Workspaces. but even has Gemini Advanced as a personal subscription.

• Need guidance or assistance tracking down threats or have logging data explained to you? Ask Gemini.
• Need help migrating the Database, building Looker Dashboards, or perhaps using Bigquery to tie in multiple datasets? Ask Gemini, - which now has the context of the project you're working on, for a more relevant contextual response.

For experienced developers, enhanced development environments with streamlined workflows, pre-built templates, and automated functionalities significantly accelerate the creation and deployment of complex AI applications.

For the typical User/Architect, Low-code/no-code solutions empower Cloud Architects and developers to getting started building Cloud infrastructure and AI applications without requiring extensive programming expertise. The recent enhancements in AI development tools have dramatically lowered the barriers to entry. This combined approach can lead to further expansion of AI-powered solutions across industries, transforming everything from customer service experiences to product development and scientific discovery.

5. Access to the vast ecosystem, transforming Ideas into production solutions. The expanded capabilities of Cloud ML/AI platforms like Vertex AI exemplify how Google is democratising AI development, giving enterprise developers at all levels the tools to transform ideas into production-ready solutions in one place. The latest update is that VertexAI allows users to deploy Models from Huggingface - a public LLM Registry with over 600k models!

This access to a comprehensive ecosystem means that enterprises can rapidly prototype, test, and deploy AI applications without the logistical pain previously involved.

Both conferences featured a long list of case studies from leading companies that have successfully integrated these technologies. For instance, a retailer using AI to customise shopping experiences, or entertainment in real-time or a healthcare provider leveraging ML to improve patient diagnostics. The cost benefits and contact centre efficiency for example in terms of customer satisfaction speaks for itself. Change: At first slow, and then sudden...

JP PoV: This is the age of unprecedented innovation and disruption at all levels - Both for Enterprise User relying on vast in-house data, and a general generativeAI user. Some levels are more acutely impacted than others. You may have the expert Python Jedi skills, and even skill, yet GenAI democratises access to such knowledge levelling up Junior developer to the near, near-equal playing field. Checkout VSCode Assist with Gemini. Consider this: project owners can ask AI to ingest vast amount of Project Updates, whether in Google Workspaces to ensure the delivery risks are flushed out and addressed earlier within the delivery sprints.

The Bottom-Up Adoption, at Full Speed

The Office Worker and General User Adoption Behaviours

Generative AI has enabled rapid thesis prototypes, codified, launched, and the ability to rework adapt and re-launch ideas as quickly as possible without breaking the proverbial bank. This does not just impact and benefit enterprises, as cloud technology with AI Tools and Services (autoML, white-label, low-code) are accessible to users, developers and businesses of all sizes.

Advanced: You can run many highly performant models entirely offline, without rate/cost limits.

Enterprise Adoption may lag after all

We are seeing the enterprise-driven technology adoption perhaps even lag behind the user adoption of GenerativeAI. This may be due to internal organisational inertia, in regards to overall AI Strategy, or effective holistic Data Management, - yet to organise, sort and clean such vast amounts of data across distinct internal organisational units and devise an acceptable user access policy, governance and the like. It will take time.

JP PoV: I believe it's a classic case of Innovator's Dilemma. The difference is that this time Enterprises are sitting on a vast amount of data, that if well organised, catalogued, cleaned and so on, can form a meaningful edge in the business marketplace compared to any new green disruptive entrant. New Market entrants may have less to lose and happy to pick off-the-shelft, fine-tune and deploy GenAI enabled service for it's customers.

Identifying a viable business use case for GenerativeAI augmentation and use-case may be the first challenge, but getting the right team together to run with the MVP, POC, (despite all odds) figuring out what quality data can be extracted and leveraged for the AI, before even considering taking any solution into production may be the second quite real GenAI Adoption snag for time being.

Your option is to skill up, fast. This may be a surprise, but with the right marketing, "Every Kevin, Jenna and their Dog are AI experts on TikTok", the reality is that there is a real skill shortage.

It's most time-sensitive to get your organisation, and team introduced to a viable and sustainable learning programme, depending on technical levels, from a High Level for effective Prompt Engineering with Generative AI, to more Machine Learning deep dive for Foundations.

• Beginners: Introduction to Generative AI via Skillsboost
• Advanced: Machine Learning Specialization
• Google Cloud Machine Learning Exam Prep Strategy Guide

More of Google Cloud Next

If you want a more thorough checklist of the announcement itinerary, do check out these blog posts already written up by the Google Team. They break down all the nitty-gritty of the conference and what you need to learn about, like;

• All 218 things we announced at Google Cloud Next ’24 – a recap
• Top Migration Announcements from Next
• 28 Startups that launched at Google Next 2024

Finally. My Take on The #AI Road Ahead

The AI revolution is here, and it's not slowing down. Check out this exponential charge about AI, Performance and likely The Singularity we're heading towards in the years ahead. It's Eye-watering.

Now, look at The DATE that was published. "We're fine, It's fine, I'm fine" meme comes to mind.

To thrive in this new era, we must embrace a culture of continuous learning and experimentation.

I hate to say this, but for those Bachelor's Degrees in the Information/Digital industry many of us have worked hard for 5+ years - They are almost entirely irrelevant now.

The Digital pioneers and consultancies are uniquely positioned to guide this journey. By staying informed, upskilling, and applying these technologies strategically, we can unlock unprecedented value for our clients and help them navigate the exciting future of AI and ML.

JP PoV: If you can be productive x10 times over. Would you sit back and do nothing with 90% of your time on your hands? Unlikely. It's more likely that you and many others will create x10 times more new solutions & value. It's The Age of Abundance

As we all collectively stand at the forefront of the AI revolution, - if anything - it's the Importance of;

• Continuous Learning and proactive adaptation - it's never been clearer. Whether you're a professional or an Enterprise, Empower yourself, Learn the fundamentals and then learn how to use these tools to be effective in your role. If you're running an organisation, employees must be upskilled and trained on Public GenAI(External) vs Enterprise AI (Internal) to reduce IP leaks. Better yet, consider investing in enabling GenAI Chatbots for Internal use - which you can ground in Enterprise data, which would make your organisation more effective and performant.
• Start innovating and experimenting with AI somewhere and start soon. If you are still struggling to see where your organisation and yourself find itself with its business strategy a year from now, - for now, you're not alone. Plan AI strategy, and find those so-called "low hanging" use cases, to prove and validate business-value-add. If your organisation is lacking skills or needs support - Reach Out, DM me. Our consulting organisation is working across many projects and is rapidly deploying PoCs that work, tailored to customer needs.
• Plan to roll out GenAI and ensure your organisation's MLOps tuning an Ethical and Responsible AI. It's not just about "Org Data as Org Moat", but how you leverage it for AI and AI guardrails in-place. Before Production rollout that is. AWS, Google and Microsoft all have Responsible AI Practices and Guides

I hope you've enjoyed this relative update on the #Cloud conference and #GenAI, AI/ML with my Poing of View (PoV) on the impact of this disruption.

Like, share, and Follow

JP

PS If your organisation needs support realising the Cloud and GenerativeAI advantages, - we're here to help. Reach out and DM.

Plus, we're always hiring great people who are keen consultant and problem solvers at heart. DM to learn more.