I have entered the world of AI cautiously and with an open mind.

The US are gripped with the AI buzz, but what I am reading in some Facebook groups is very worrying, so needs to be researched carefully.

‘What AI can I use to scrape websites?‘

Today I saw and AI image of Sheikh Maktoum of UAE in and AI generated image standing with a random man talking about the news that the UAE is giving free access to Chatgpt ( Sheikh Maktoum is no longer the President of UAE anyway ). I asked about gaining consent to use his image but they just don’t care about consent. People are using images of real people and creating AI videos and graphics.

When Google launched AI Overviews it created a completely new way for users to search in Google by offering more in-depth answers to questions, so the questions and answers became more complex.

Google views AI Overviews as one of their

‘most successful Search launches in the past decade‘.

Google believes people are starting to use Google AI search very differently and that coupled with the speed of delivery anticipates a huge growth in this area.

Google are aiming to go beyond information gathering to information intelligence. Google finds the information from websites so they will need to be factual.

Google are rolling out AI Mode across the US in its latest AI search project.

AI Mode increasing search capability by offering advanced reasoning and ability to offer follow-up questions and links to relevant websites and resources.

AI Mode uses a query fan-out technique breaking down the question into sub questions then diving deep into the world wide web to find specific answers to the queries.

Google will also use Gemini’s capabilities.

Google Lens is a visual search tool but now they are combing Project Astras live capabilities into Search.

Search Live is a feature that lets users have real-time conversations with Search using their smartphone camera.

If you need practical help then you can tap the ‘Live’ button in AI Mode or in Lens, point your camera and ask a questions.

The search becomes a classroom! with links to resources i.e. videos and more

Project Mariner – observes, plans, actions

To help visitors with tasks such as booking tickets, project Mariner will be added to AI Mode.

Asking questions such as “Find 2 cheap tickets to Lisbon for Tuesday 9th September from Gatwick” and the AI Mode will deep search to analyse all available options.

AI Mode with present best options for that question on websites and you can select the one you want to go with.

Google will partner with all the ticket agents such as Ticket Master to get the best prices and offer the best experience.

Google will track you using cookies to then offer you a more customized experience.

Google AI will get to know you so well it will even know where you sat at the last concert you went to, where you got your hair cut and what bank you bank with.

Cookies track you to get to know your shopping habits, but with AI, they will be able to do a lot more with your data.

AI Mode will also help transform numbers to charts. AI Mode will be able to create visuals from datasets.

Create analysis for all sorts of topics from sport to farming.

AI Mode will be rolled out to the US in May 2025 with plans to roll out further.

Agentic AI refers to a type of artificial intelligence that operates autonomously, making decisions and performing actions without constant human intervention.

It’s characterized by its ability to learn, adapt, and pursue specific goals in dynamic environments. Agentic AI systems can be composed of multiple AI agents working together to achieve complex objectives.

Key Characteristics of Agentic AI:

Below is a list of characteristcs  of Agentic AI

Autonomy: Designed to act independently, making decisions and taking actions based on their own understanding of the situation.

Adaptability: They can adjust behavior and strategies in response to changing circumstances and new information.

Goal-driven: They are given a set of goals that they work with to aim to achieve.

Learning: They can learn and improve performance over time.

Complex reasoning: Leverages complex reasoning and planning to solve multi-step problems and achieve complex objectives.

Probabilistic and adaptable: They use probabilistic reasoning and adaptability to handle dynamic environments and make decisions based on patterns and likelihoods.

How Agentic AI Works:

The Agentic AI typically involves the following:

Gathers data from the environment.

Processing data to understand the situation.

Making decisions and taking actions based on understanding.

Improving and evolving based on feedback and experience.

Examples of Agentic AI in Action:

AI chatbots: Used to as customer support.

Personal health assistants: Monitoring health, creating health plans and predicting future health issues.

Autonomous vehicles: Used to monitor traffic flow.

Robotic process automation (RPA): Automating tasking such as data entry.

Supply chain management: Optimizing the supply chain process, predicting demand.

Cybersecurity: Identifying threats and taking action.

The Benefits of Agentic AI:

It is very clear that AI increases efficiency by automating some tasks and processes which should provide businesses with cost cutting and less mistakes.

AI can be used for decision making as it can read and analyze large volumes of data and make programmed decisions on the data output.

Productivity is going to be faster and more accurate using AI. It frees up time for other tasks.

AI is more adapablt to change than many humans, so it is able to adapt to changing conditions quickly and decisively.

Challenges of Agentic AI:

The autonomy of AI raises ethical questions about accountability and bias.

Building AI to work accurately and perform tasks efficiently will require expert programming and understanding.

AI often rely on large volumes of data, so that data needs to be collected and needs to be accurate.

Privacy around the data gathering is still questionable.

Answer Engine Optimisation (AEO)

Companies will need to know how to ethically navigate AEO, how to incorporate targeted ads, create content that answers those important questions and possibly change the traditional nature of relationships with customers.

Consumers need to understand the AI space in terms of protecting their personal data and privacy.

Microsoft Copilot Studio

I have just discovered Microsoft Copilot and my life is now complete. It will give me the answers to all my burning questions.

Microsoft Designer

an AI-powered graphic design and image editing tool that allows users to create and edit designs and images with ease.

Find out more

OpenAI announced the acquisition of Jony Ive’s design company that’s building a breakthrough AI-first device. While they didn’t share any previews or visuals, Sam Altman raved about the prototype he’s seen and said it’s the best thing device ever seen!

Anthropic launched Claude Opus 4, its most intelligent LLM (Large Language Model). But before they did, they had to fix a few issues including the chatbot’s ability to lock people out of accounts, develop self-propagating viruses and write notes to its ‘future AI self’.

As the USA speeds along in the AI space, how do we in the UK keep up and do we need to?

There is a lot of AI noise on my socials, making it difficult to know where to start, it is overwhelming but I want to focus on how AI can help businesses with their operational success.

The chatbots we have to talk to on websites are quite frankly, annoying at best.

To get a grip with this new AI space we need to pick areas of the business we can adopt AI to i.e. Lead generation and integrate that into our websites.

Launching the Artificial Intelligence Playbook for the UK Government

The AI playbook sets up corporate guidance for policy professionals and decisions makers and technical advise on how to buy and implement AI solutions.

There are 5 key points.

10 principles

Key principles that people should follow when using AI in government.

Introducing AI

The foundations of AI (including generative AI), its applications, capabilities and limitations.

Building AI

Designed for a variety of roles working on AI projects and contains both corporate and technological guidance.

Advice on understanding if AI is the right tool for the job, what use cases should be avoided, how to conduct user research on AI, how to build an AI team and acquire the necessary skills, and how to buy, procure and implement AI products.

Using AI safely and responsibly

The legal, ethical, security and governance aspects of AI projects that team leaders and senior decision makers should consider when working on AI, discussing topics such as bias, privacy, copyright, data protection, and cybersecurity.

Appendix with AI use cases in the public sector:

Contains case studies on the development of several AI and generative AI solutions.

AI requires Data, so who and how do tech companies gather the data ethically and what are their legal obligations around obtaining and storing this data?

Will trust be erroded by the use of AI if consumers data is exposed and how do Governments tackle this?

Here are four strategies to get you started:

Privacy by Design

Begin by identifying the minimum types of data your AI system needs and what you’ll use it for.

Safety and security

Develop a system to remove personally identifiable information and safeguard people’s privacy during training and deployment.

Get clear and specific consent

Adopt privacy-led marketing and make sure you’re upfront and your needs are expressed in easy to understand language.

Audit and adjust

Conduct regular reviews of your system’s use of data and make adjustments to keep customer privacy front and center.

To source AI requires huge power inputs and the extraction of resources for all the components.

The data, where is it coming from and do we have consent to take it? Is the data accurate, verifyable?

What is the law around AI? Are we keeping up.

How do Governments and businesses adopt the benefits of AI whilst protecting consumers data and privacy? This is a machine that feels like it could get out of control and leave people behind.

In short, in the coming years, there is barely an aspect of our society that will remain untouched by this force of change. But this government will not sit back passively and wait for change to come. It is our responsibility to harness it and make it work for working people. And it is our responsibility to make sure that Britain maintains its position as a world leader in AI, even as the competition increases;

Sir Keir Starmer

In Summary

There is a lot to take onboard, a lot to digest and lots of pros and cons. My worry is that people might get left behind if they don’t have the skills, or the equipment required to use AI. We need intelligence to set up and create workable systems and engineers to fix it when it breaks.

Small businesses can take advantage of AI lead generation and automation into their business practice.

Digital Marketers will be able to assist businesses implement the right tools to fit business needs.

The best recommendation I can make is to find people you trust to recommend the right AI tools and people that can manageme them.

 Read the UK Government plans for AI