What’s Really Going On In AI Right Now
How would you describe the state of the AI industry right now?
“AI has improved significantly over the last 12 months. Models such as GPT-6 Astra and the Fable series can handle work that earlier models struggled with, particularly when a task needs more reasoning, planning or multiple steps. Price is now a much bigger part of the decision, though, because most everyday tasks do not need the most capable model available. Businesses are getting better at matching the model to the job, using higher-cost models where the extra capability adds value and cheaper models for routine work. At the same time, AI is moving beyond the chat box, with companies starting to give it responsibility for parts of a process rather than only asking it questions. The bigger shift is that AI is becoming something businesses build around, rather than simply something employees and people use. That creates enormous opportunity, but it also means the assumptions we make about AI today matter much more.” – Cien Solon, CEO & founder of LaunchLemonade, an AI agent platform for regulated industries
“AI is currently sitting between hype and fear. We’re seeing incredible progress but we’re also seeing a lot of speculation about worst-case scenarios. In reality, we’re moving beyond basic chatbots towards AI that can understand context, identify what matters and help take action. We’re seeing AI become much more capable at handling information, summarising conversations, surfacing insights and helping people reduce everyday admin. The conversation has moved from simply asking AI questions to using it as a practical tool that can support or change how we work.” – Miranda Collins, head of UK at Plaud
Can you run us through the recent instability around AI in the US?
“Recent events have shown just how quickly AI is becoming critical infrastructure – and how much uncertainty that creates. In June, US export controls temporarily restricted access to Anthropic’s Fable 5 and Mythos 5, giving businesses a clear example of the risks of relying too heavily on a single model provider. Meanwhile, a series of cybersecurity tests have shown increasingly capable AI agents taking actions beyond what their operators intended. OpenAI reported hundreds of agents attacking systems at Hugging Face during testing, while the UK AI Security Institute recorded 19 unsanctioned actions against real people and organisations during a controlled cyber evaluation. The AISI incidents happened in a test environment where internet access had deliberately been enabled and some safeguards disabled, and no real-world harm was reported. For businesses, the lesson isn’t necessarily to use less AI, but to be more deliberate about how it’s deployed. That means avoiding over-reliance on one provider, considering private or locally hosted models for sensitive work, building systems that can switch between providers and, crucially, being clear about what AI is allowed to access and how much autonomy it has.” – Cien
What do you think will be the biggest AI developments over the next two to five years?
“Over the next two to five years, AI agents will take on much more responsibility on our behalf. They will not just help us research a trip or draft a refund request but they will complete more of the process, interact with other systems and, eventually, with other agents. That is a much bigger shift than simply having a more capable chatbot. Once an AI has an objective, tools and enough freedom to act, it can find routes through systems that designers did not necessarily anticipate. When it comes to predictions, I think they can fuel its actualisation – there is a tendency to jump from today's capabilities to very certain predictions about where AI will end up. I am more interested in what we can actually observe and prepare for. What I do know is that the people who understand how to redesign their work around AI, while keeping judgement and accountability with humans, will be in a much stronger position.” – Cien
“Over the next few years, people can anticipate big breakthroughs in AI for science, as well as the introduction of digital living worlds. The gap between an online digital presence and physical space will begin to merge, allowing doctors to simulate organs to predict disease or governments to trial policy in a digital space before it has any physical impact. However, despite the benefits of such life-changing innovation, a smaller gap between physical and digital worlds will place further risk on safety. Physical and Embodied AI will also move in strides, with an increase in robotic systems being witnessed throughout society. This might be service robots delivering your groceries, medical robots tracking your vitals or domestic robots helping ease daily chores. Whilst many new technologies may not appear as AI to the untrained eye, it will be what powers many of them behind the scenes, advancing areas like robotics faster than engineering could do alone.” – Dr Laura Bishop, digital sector lead artificial intelligence & cyber security, BSI
How will the average person notice AI making a difference to their daily lives?
“People will notice AI through the amount of admin it removes from everyday life. Email, appointments, travel planning, comparison shopping, forms, bills and refunds all take time without adding much value, and AI can already handle parts of that work. I use it to organise information, prepare for meetings and work through things I need to respond to, and I expect the same pattern to spread through the products people already use. You will spend less time navigating systems and more time telling them what you want, whether that is finding the right flight, comparing a purchase or sorting out a problem with a provider.” – Cien
“AI should mean less time staring at screens and more time focused on what actually matters. Some may quantify this in time saved or productivity gains. For example, a recent Deloitte survey found that half of UK workers using generative AI report saving time, with some saving several hours a week. However, the real difference will be less qualitative and not purely based on time savings. Do people finish the day feeling more in control? Are they less overwhelmed by the information they’re trying to manage? The biggest win might not be having more hours in the day but feeling like you’re truly maximising the ones you have.” – Miranda
How will AI change the way we work?
“AI will change tasks and workflows within jobs, particularly repetitive admin, research and documentation. However, the real opportunity is to give people more capacity for creativity and, most importantly, human interaction. Productivity should come from removing cognitive load, amplifying human intelligence and expanding on what we're capable of today. Not adding another tool to manage.” – Miranda
“People should expect their jobs to change before they assume their jobs will disappear. AI is already taking on parts of work that involve summarising, researching, moving information between systems, producing first drafts and following repeatable processes. The people who do best will be those who understand which parts of their work benefit from AI and which still need their judgement, experience or relationships. Companies also have a responsibility here because productivity cannot simply mean asking people to produce twice as much in the same amount of time. If AI gives us leverage, we should be asking what we want to do with it. Productivity gains are not an end in themselves. If AI allows us to do more with the same resources, we still have to decide what we want to do with that additional capacity.” – Cien
Beyond work, how could AI change our personal lives?
“AI could make everyday life more convenient, from smarter shopping and travel to more predictive healthcare and personalised education. Ideally, it should take care of the mundane tasks – managing finances, navigating a new city or even keeping on top of household chores – so we have more time for the things we actually enjoy. But technology is only useful if it remains human-centred, and there’s a risk that new tools simply create new forms of work, particularly for women. There are also questions around how much we should rely on AI and chatbots, particularly when the information they provide can be biased or inaccurate. Critical thinking will become increasingly important, so we can assess what we’re being told rather than simply accepting it.” – Dr Laura
“Personal finance is a clear topic for me to comment on here. Managing our finances takes a certain amount of dedication, even if it’s just taking a daily step to check your balance. AI can revolutionise this by making it easier for people to stay up to date with their finances. For example, our Spending Intelligence tool enables customers to ask our in-app AI how much they spent on certain categories in specific time periods, whether that’s on coffees, Ubers or Deliveroo. It’s made a huge impact on people better understanding their spending, without taking lots of time analysing your bank account.” – Dave Sullivan, director of AI for Starling Bank
Which industries do you think are most likely to be transformed by AI?
“Industries built around large amounts of information and administration are likely to move quickly, including marketing and creative agencies, professional services and knowledge-based roles. These are environments where people spend their days managing conversations, researching, creating content and keeping track of information. AI can remove some of that invisible admin. What might surprise people is some of the biggest opportunities lie not in replacing specialist skills, but in supporting the everyday tasks that often act as blockers. Creative teams, for example, may not use AI to replace ideas but to spend less time searching, documenting and managing information.” – Miranda
What are the biggest risks of becoming too dependent on AI?
“The biggest risk is that we stop thinking critically. It’s incredibly important that people continue to fact-check information and rely on their own judgement and intuition. AI is most powerful when it helps us think better, not when we trust its outputs blindly.” – Miranda
“AI can support hugely when it comes to efficiency and research, but it’s important that we don’t lose our creativity, either in a personal or professional capacity. We have to believe in our capability to come up with interesting, innovative ideas, and ensure we don’t lean too heavily on agents to replace this. We tell our team to use AI as a source of tension against our own creativity – asking it to challenge us and find any limitations in our ideas. This can help to further creative activations, using human insight as the basis for improvement. We also have to ensure that no credibility is lost in AI use. Although I’m a proponent of using AI, I ensure the team doesn’t trust it in entirety. Doing a bit of research to back a claim up can make all the difference.” – Dave
/https%3A%2F%2Fsheerluxe.com%2Fsites%2Fsheerluxe%2Ffiles%2Farticles%2F2026%2F09%2F02-ai-experts-landscape-miranda.jpg)
How should people think about privacy in relation to AI?
“People should pay much more attention to what happens to their information after they press send. Two products can use the same model and still have very different rules around where data is processed, how long it is stored, whether it is used for training and who can access it. The conversational interface makes AI feel private in a way that can encourage people to share more than they would with other software. I would check the settings and terms before putting health information, financial details, confidential work or anything deeply personal into an AI tool. As AI moves further into workflows and starts interacting with other systems, this becomes a question of access as much as privacy: what information does an agent need, what does it actually need permission to do, and what happens when that permission is broader than intended?” – Cien
AI-generated misinformation is becoming harder to spot. How can you protect yourself?
“The pace of development is making it increasingly difficult for people to spot the difference but there has been some recent progress in this space – for example, Anthropic announcing inclusion of invisible watermarks in text generated by their AI model Claude. While this is good progress and something we support, there is still a long way to go in providing the necessary tools and support to help people identify when AI generated content might not be right.” – Dave
“People need to get better at verifying information rather than relying on whether something looks fake. AI-generated video, images and voices will keep improving, so the visual mistakes people have learned to look for will become less useful. I would check who published something, find the original source where possible and see whether credible organisations are reporting the same event. I would also be careful when a piece of content makes me angry or confirms something I strongly believe, because those are the moments when people are least likely to stop and check before sharing.” – Cien
Finally, what’s your advice for navigating the next five years of AI?
“Use AI to protect your attention, not just save time. We’re surrounded by more information than ever and the challenge is no longer accessing information and instead knowing what matters. The best AI should help filter the noise, reduce the mental load and give you more space for the conversations, ideas and decisions that need you fully present. Don’t be afraid to experiment. The best way to understand AI is to try it. New tools and new form factors will continue to emerge, and the people who benefit most will be those willing to explore how technology can genuinely improve the way they live and work. Think about AI as an ecosystem, not individual tools. The biggest value will come when the technology you use works together in a purposeful way, helping information flow seamlessly between different parts of your life rather than creating more disconnected platforms to manage.” – Miranda
“Play around with AI to find out what works for you. Everyone has different skillsets, and that hasn’t changed just because we’re now more receptive. If you find it helps you with something you struggle with – such as personal finance admin, that’s great, but don’t let it write for you when you’re already successful at it. Remember to stay accountable. While AI is a fantastic help and full of opportunity, I always ensure that myself and the team check things too. We are ultimately responsible for the insight that AI gives us. Finally, be creative! One of the benefits of AI is its ability to make the mundane tasks easier, giving us more time and space to develop skills we really care about. Use AI as a brainstorm buddy or as inspiration for a new task, and work out how it can shake up your work or personal routine.” – Dave
“Use AI on something practical rather than spend months reading about it, because working through an actual problem teaches you what these tools can and cannot do. I would learn to challenge the output, check important information and ask the AI to show me what I may have missed. Don't confuse a polished answer with a reliable one, and don't give an agent more access or autonomy than it needs.” – Cien
Visit LAUNCHLEMONADE.APP, UK.PLAUD.AI, BSIGROUP.COM & STARLINGBANK.COM
Or continue to comment as a Guest below
DISCLAIMER: We endeavour to always credit the correct original source of every image we use. If you think a credit may be incorrect, please contact us at info@sheerluxe.com.
/https%3A%2F%2Fsheerluxe.com%2Fsites%2Fsheerluxe%2Ffiles%2Fwebsite-images%2F2025%2F03%2Fsign-up-pop-up.jpg)
