AI For Real

Author: AI For Real Team

  • AI is creating a new kind of startup: one that gets smaller as it grows

    AI is creating a new kind of startup: one that gets smaller as it grows

    The global startup world has long followed a simple formula: raise money, hire people, build faster and increase headcount as revenue grows, but artificial intelligence (AI)is beginning to turn that model on its head.

    The trend extends beyond individual companies, with LinkedIn data showing that US startups founded in 2022 with at least five employees averaged 20.7 workers by their third year, compared with 32.5 for similarly sized companies founded in 2018.

    Data from Carta points in the same direction, showing average headcount at Series B startups falling from 53 employees in 2023 to 45 in 2025, while average Series D headcount fell 29% from its 2023 peak.

    The phenomenon is not confined to the US, with an AWS study of more than 3,400 founders across 20 countries finding that AI-native startups are reaching billion-dollar valuations in an average 3.5 years, with half the staff required by comparable businesses before generative AI emerged.

    Canada offers further examples.

    The result could be a profound change in what it means to build a company, as founders increasingly ask whether a new employee is necessary at all when an AI agent can perform the work, leaving humans to concentrate on judgement, relationships and decisions that remain harder to automate.

  • What is a tool call? How AI actually does things

    What is a tool call? How AI actually does things

    In our ‘AI Primer’ segment today, we will tell you about “tool call”.

    When you ask an AI model a question, it can generate an answer from what it has learned, but a tool call allows it to go beyond its own knowledge and interact with an external system.

    A tool call is essentially a request from an AI model to use a specific tool, such as a web search, calculator, database, computer program or application programming interface.

  • Mistral and Mozilla develop AI assistant designed to protect user privacy

    Mistral and Mozilla develop AI assistant designed to protect user privacy

    Mistral and Mozilla are joining forces to develop an artificial intelligence (AI) assistant for Firefox that aims to give users the benefits of AI without requiring them to hand over their personal information.

    The companies said their partnership is designed around privacy, with Mistral committing to a model in which users’ conversations with the assistant are not retained or used to build profiles about them.

    Why this?

    The technology will power Mozilla’s experimental Smart Window, an AI feature designed to help people search the web, understand information and work with content across multiple browser tabs.

    Unlike many online AI services, the smart assistant intends to keep users’ interactions private, reducing the amount of personal data that needs to leave the browser or be stored by the companies providing the service.

    For users, this means they could ask an AI assistant to summarise information, answer questions or help navigate the web without necessarily creating a lasting record of those interactions.

    The partnership reflects a broader concern about the amount of personal information that AI services can collect as they become more deeply integrated into everyday online activities.

    As some of you may know, Mozilla, the organisation behind Firefox, has long positioned privacy and user control as important principles, while French AI company Mistral has developed models that it says can be deployed with greater control over data.

    The companies said the partnership will combine Mozilla’s experience in building privacy-focused products with Mistral’s AI technology to explore a different approach to AI-powered browsing.

    Smart Window is being introduced experimentally, allowing Mozilla and Mistral to test how people use an AI assistant while addressing concerns about privacy and control.

    The initiative therefore goes beyond adding another chatbot to a web browser, as the companies are testing whether AI can become part of everyday browsing while keeping users’ personal data under tighter control.

    Image credit: Mistral/Mozilla

  • Synthetic intimacy: When People Form Emotional Bonds With AI

    Synthetic intimacy: When People Form Emotional Bonds With AI

    Synthetic intimacy describes the emotional closeness people develop with artificial intelligence systems that are designed to simulate conversation, companionship, empathy and personal attention.

    The term has gained attention as AI assistants and companion apps become more capable of remembering conversations, adapting their responses and maintaining a consistent personality across interactions.

    Unlike traditional search engines or productivity software, these systems can create the impression of an ongoing relationship by responding conversationally, asking follow-up questions and appearing to recognise a user’s preferences and emotional state.

    For some users, that interaction can become a source of companionship, particularly when AI is available at any hour and does not display the social friction associated with human relationships.

    AI companies are also increasingly designing systems to feel more natural, with voice conversations, personalised responses and persistent memory making interactions resemble relationships rather than conventional software use.

    That has raised questions about where useful companionship ends and emotional dependence begins, particularly when users attribute intentions, feelings or understanding to systems that do not experience emotions in the human sense.

    Researchers have also examined whether highly personalised AI could encourage people to disclose sensitive information or spend more time interacting with systems that are optimised for engagement.

    The commercial implications could be significant because synthetic intimacy creates a potential market for AI companions that goes beyond conventional productivity and entertainment applications.

    At the same time, the technology raises questions about transparency, consent and the responsibilities of companies when users form strong emotional attachments to their products.

    The phrase therefore captures a broader shift in human-computer interaction, from people using AI primarily as a tool towards people treating increasingly conversational systems as social companions.

  • Tilly Norwood Interview Puts AI Actors Back In Hollywood Spotlight

    Tilly Norwood Interview Puts AI Actors Back In Hollywood Spotlight

    An AI actor giving a press interview??!!

    Yes, it just happened.

    Tilly Norwood, the AI-generated actress at the centre of a growing Hollywood debate, has reignited controversy after appearing in an interview with NBC’s “Today” programme.

    Norwood was questioned by NBC entertainment correspondent Chloe Melas about the backlash surrounding her creation and fears that artificial intelligence (AI) could eventually replace human actors, writers and other creative workers.

    The interview has created a buzz because Tilly is no longer simply an experimental AI character appearing in social media clips, but is increasingly being presented as a potential performer with a place in the entertainment industry.

    Tilly insisted that her purpose was to act as a creative tool rather than a threat to human performers, pushing back against the idea that her existence necessarily means fewer opportunities for actors.

    The controversy is particularly sharp because Tilly has already been positioned for a feature-film role in “Misaligned”, a project being developed by Particle6, the company founded by comedian and former actor Eline van der Velden.

    Van der Velden has said the intention is not to put Tilly into conventional live-action productions, but to develop AI-native films in which synthetic performers can operate alongside AI-generated environments and stories.

    Human Performers Worried

    Tilly has also argued that audiences should see AI as another creative medium, with her creator previously comparing the character to a new kind of paintbrush rather than a replacement for an actor.

    But that reassurance has done little to settle the debate because the technology is developing precisely when performers are already worried about their jobs, likenesses and bargaining power.

    The interview therefore touches a much bigger question than whether audiences like Tilly Norwood: it asks whether a computer-generated performer can eventually become a genuine entertainment star.

    That question has become harder to dismiss as Particle6 develops “Misaligned” and says it is building a wider roster of AI-generated performers, while acknowledging that the technology could drive a major transition in creative jobs.

    For now, Tilly remains a synthetic character rather than a human actor, but her growing media profile shows why she has become such a powerful symbol of Hollywood’s argument over AI.

    The NBC appearance gives that debate another high-profile platform, and the uncomfortable irony is that an artificial actress is generating very real anxiety among the people whose jobs she is designed not to replace.

    Image credit: NBC

  • How Real-Time AI Is Rewiring ‘Live’ Sports, TV, And Radio Broadcasts

    How Real-Time AI Is Rewiring ‘Live’ Sports, TV, And Radio Broadcasts

    If you have ever watched a football match or a ‘live’ television broadcast, you know that time is a strict tyrant. In the traditional world of television, what happens on the screen is a high-stakes magic trick that requires hundreds of technicians, miles of cables, and split-second human reflexes to pull off.

    But behind the scenes, the foundation of this industry is undergoing a massive earthquake. Artificial intelligence (AI) has officially moved out of the post-production editing room and onto the ‘live’ stage, operating with the speed and precision of a human reflex nerve.

    Think of traditional ‘live’ broadcasting like a massive, old-fashioned railway system. Every piece of cargo, from camera angles to commentary tracks, has to be manually switched, routed, and checked by human operators at every junction. If a broadcaster wanted to translate a ‘live’ soccer match into five different languages with native emotional weight, they essentially had to build five parallel railway tracks, hiring separate crews and commentators for each.

    But these days, real-time AI is acting as an intelligent, lightning-fast digital switchboard that handles these complex logistics on the fly, transforming a single ‘live’ feed into a localized, hyper-personalized experience for millions of viewers worldwide.

    Click here to read the rest of the post.

  • Auditing Artificial Intelligence Bias: A Student Guide

    Auditing Artificial Intelligence Bias: A Student Guide

    As artificial intelligence (AI) becomes deeply integrated into everyday academics, students find themselves uniquely positioned to evaluate its outputs critically.

    Young learners routinely interact with machine learning models for research, writing assistance, and creative projects. Because these systems are trained on vast datasets reflecting historical human behavior, they often perpetuate hidden prejudices and stereotypes.

    Learning to audit artificial intelligence bias is an essential modern skill that empowers students to use technology responsibly and demand greater accountability from developers.

  • AWS Gives Students Free Access To Kiro For A Year

    AWS Gives Students Free Access To Kiro For A Year

    Amazon Web Services (AWS) has announced the expansion of its free Kiro program, giving students at 132 universities across 18 countries access to the AI-powered software engineering agent for one year.

    The initiative aims to help students develop practical artificial intelligence (AI) and software-development skills before entering the workforce.

    Key Features

    Under the program, eligible students receive 1,000 Kiro credits every month without requiring a credit card. They also gain access to the latest AI models and Kiro Web, a zero-install version of the tool.

    This expanded access is intended to overcome the limitations of conventional free AI tools, which often impose usage restrictions that can interrupt student projects.

    Kiro uses a “spec-driven development” approach. Instead of simply generating code from prompts, students first define requirements, design decisions, and implementation steps. Kiro then helps generate code while students review the results and ensure they meet the original specifications. This encourages critical thinking, problem-solving, and responsible use of AI alongside technical skills.

    Results and Global Expansion

    The program began as a pilot in March 2026 involving 11 universities in two countries. Thousands of students subsequently used Kiro to develop projects, including an accessibility application for visually impaired users and a voice-enabled wellness platform connecting patients with medical professionals.

    The latest expansion increases university participation twelvefold and extends the initiative across Asia Pacific, Europe, Latin America, and North America. Participating institutions include universities such as the National University of Singapore, Imperial College London, the University of Toronto, and the Massachusetts Institute of Technology.

    Image credit: AWS

  • Cognizant’s AI Hiring Push Shows How New Jobs Could Emerge As Old Ones Change

    Cognizant’s AI Hiring Push Shows How New Jobs Could Emerge As Old Ones Change

    For years, one of the biggest fears surrounding artificial intelligence (AI) has been simple: if machines can do more of the work, what will people be left to do?

    A recent move by Cognizant offers a different way of looking at the question.

    The technology services company says it plans to hire 1,500 US college graduates and expand two new roles, called Frontier Certified Engineers and Frontier Business Operators, to a combined 15,000 people.

    These are jobs that are emerging specifically because companies are trying to put AI to work.

    The example is important because it shows that AI does not only replace existing tasks. It can also create demand for people who know how to build, manage, check and apply AI systems.

    As businesses adopt these tools, they need employees who understand both the technology and the business problems it is supposed to solve.

    Cognizant says one of its teams recently redesigned an account-management process for a food-service company using 17 AI agents, saving account managers around 11 hours a week.

    That does not necessarily mean the account managers became unnecessary. Instead, some of their routine work was taken over by AI, potentially leaving more time for work that requires judgement, communication and decision-making.

    But there is another side to the story. A company announcing thousands of new AI-related positions does not prove that AI will create enough jobs to replace every job it disrupts.

    Some roles will almost certainly shrink or disappear, particularly where work consists largely of repetitive and predictable tasks. And not everyone whose job changes will automatically have the skills needed for the new positions.

    That is why the more useful question for graduates may not be whether AI will take jobs or create jobs. It is how quickly the nature of work will change, and whether education and training can keep up.

    Cognizant says it has already trained more than one million people through its global skilling programme and now aims to reach two million by 2030.

    For graduates entering the workforce, the lesson is neither to fear AI nor to assume it will solve the jobs problem. The safer bet is to understand how AI works, learn how to use it, and develop the human skills that remain difficult to automate.

    What do you have to say about this? Do write in the “Comments” section.

  • Prompt Goblin: The Person Who Can’t Stop Tweaking AI

    Prompt Goblin: The Person Who Can’t Stop Tweaking AI

    A prompt goblin is someone who treats AI prompting less like a skill and more like an obsession.

    They constantly tweak, remix, test, and over-engineer prompts to squeeze a little more intelligence out of an AI model. Give them a simple question, and they might respond with a 47-line system prompt, five constraints, three examples, and a request for the AI to “think like a world-class expert.”

    The term is playful, but it captures a real behaviour emerging around generative AI. Prompt goblins enjoy discovering the strange ways wording, context, instructions, and formatting can change an AI’s output. They experiment endlessly, often saving successful prompts like treasured recipes.

    In 2026, the prompt goblin is evolving. As AI systems become better at understanding natural language, obsessive users are shifting from clever prompts toward context engineering, agents, tools, and workflows.

    In other words: the prompt goblin may be disappearing, but probably not before trying 37 more prompts.