A year is now an awfully long period of time. The sheer speed of AI adoption has expedited all other tech cycles, making it difficult to predict what’s coming our way in the next few weeks.As a result, we are living in exciting times, with new products and updates shipping every few days, opening up opportunities and challenges we hadn’t even envisaged the week before.Despite all this buzz, however, the fact remains that AI has yet to bring in the kind of efficiencies it once promised, raising fears that we are all just riding another bubble inflated by a lot of hype.Despite these uncertainties, most of what we said in this column a year ago has been correct, though with varying degrees of success. AI is now being deployed across industries and by individual users for practical use cases. However, autonomous AI agents, as well as the use of the technology for creating new hardware and simplifying dashboards, is taking time.Yearender 2025 | Yearender 2025 | Diplomacy: After a year of rupture, what 2026 holds for New DelhiBut we are clearly on that path and the drag is mostly from adaptability and adoption. The social media flux is continuing with users becoming spectators rather than active participants, though we haven’t seen the emergence of any new platform that captures user interest.This gives us the confidence to come up with ideas that could come true in 2026. In fact, many of the predictions from last year could actually materialise now. And, not surprisingly, there is a common theme this time for all the predictions — and it needs just human intelligence to figure this one out.A new homepage for the internetThere is a drastic change happening in how we access knowledge on the internet. If Google gave us the ability to search for and find links in a list format, AI platforms are making it more natural to consume the same information in concise textual nuggets. While this has resulted in the “zero click” panic for online portals, the fact is that for users, AI formats make more sense as they deliver what they want in a personalised format and with the ability to ask follow-up questions and fill knowledge gaps.Story continues below this adThis change in user behaviour is transforming the business models the internet is based on. If click-through rate, or CTR, was the metric on which web success was measured, Generative AI has taken away the need to click on anything to access information.In the new year, expect this transition to accelerate with the widespread adoption of AI browsers like ChatGPT’s Atlas and Perplexity’s Comet. These are essentially browsers serving what you seek without the need for an additional click. They can even surf the internet for you, giving you daily updates from the websites you love without you ever returning to these domains.All of this is already threatening Google, which has had a monopoly on users of the internet, especially in countries such as India where it essentially serves as the gateway to the world wide web.If people stop entering search queries on Google surfaces, the most powerful business model in the world will be challenged. But it is unlikely Mountain View will let go of its stranglehold on the web. It is already offering AI overviews on most searches, giving users the knowledge they need without requiring them to click further. In 2026, expect this to become the norm in how we access the internet — pulling knowledge from websites rather than pushing users in for content.Story continues below this adAI-powered devicesThe browser is the most used app on any smart device. In the new year, expect smartphone and laptop companies to leverage this by making an AI browser the core of their device. AI is already a buzzword for these devices, but these interventions are superficial at best and based on an earlier understanding of how users will interact with GenAI surfaces. The fact is that not everyone wants to keep creating AI-generated images or summarising their emails. They want to do more, and maybe they want the operating system itself to do more for them. An AI-powered device can be like your personal butler — it will summarise your day for you, make appointments, and remind you of important calls and emails that need responses. Don’t be surprised if OpenAI or Anthropic announced an AI phone.If smartphones spawned the age of apps, GenAI platforms are going to take us into a post-app era by doing everything without you having to open specific applications. These app integrations are already happening, and you will only see this increase in the coming months. The interaction with your phone might be more like a conversation and may not even need a textual interface. You might be able to put away your device in your pocket as you use either goggles or earbuds to interact with non-visual content most of the time.A rewind on your lifeAn extension of these devices will be their ability to record everything you have seen and let you go back to what you might not be able to remember. If the new versionGoogle’s Android XR Glasses are any indicator, these devices are closer than we think. They will make notes from everything you see and every conversation you have, using AI to take actionable next steps so that you don’t forget anything. There will be obvious privacy issues with technology that is logging even your offline life, but there will be obvious advantages, too. And this will also be technology that is more passive than what we are used to, working in the background as we go about our daily lives. The possibilities of this new kind of AI assistant will be endless, and soon you will see the co-pilot aspect permeate into physical life as well. AI glasses will be able to tell you how to react to situations, read languages you have never learned, and solve problems you have never encountered.Story continues below this adAI agent orchestrationIf 2025 was the year we all started experimenting with AI agents, 2026 will give us the power to orchestrate multiple agents to complete complex tasks, making decisions when they need to and moving on when they can’t. Platforms that let you orchestrate AI agents, often from different companies, will gain popularity in the coming year, especially among more evolved AI users. This multi-agent world will be especially beneficial in bringing efficiencies to businesses that have to manage multiple departments and vendors, resulting in different types of data and dashboards. Deloitte’s market estimates suggest the autonomous AI agent market could reach $8.5 billion by 2026.Physical AI Those who have sat in an autonomous Waymo taxi in San Francisco may think it is a better, safer ride than Uber. And this lack of human intervention, or distraction, could drive up the adoption of autonomous public transport across the world, followed by drone deliveries, especially to places humans can’t easily reach. As long as they have a good power supply, these autonomous automobiles can manage on their own without getting stuck on sidewalks or running into lamp posts. This newfound autonomy is also going to result in the adoption of robots from companies like Tesla, Boston Dynamics and Figure AI that do everything from folding and ironing clothes to taking your dog out for a walk and, of course, increasing productivity and efficiency in factories. However, this will happen only if their per unit costs come down to under $50,000 from where it is in the prototype stage now.