AI For Real

Author: AI For Real Team

  • Just 10 minutes of AI assistance can make people less willing to tackle difficult tasks, study finds

    Just 10 minutes of AI assistance can make people less willing to tackle difficult tasks, study finds

    Using artificial intelligence (AI) tools such as ChatGPT to solve problems may make people less willing to persevere when the technology is unavailable, according to new research involving 1,222 participants.

    The study, co-authored by Brian Christian, a research fellow at the University of California, Berkeley’s Center for Human-Compatible AI, found that people who relied on AI assistance struggled more with independent problem-solving after the tool was withdrawn, raising questions about the long-term effects of using chatbots for everyday tasks.

    Researchers conducted three randomised controlled experiments involving participants who tackled maths problems and reading comprehension exercises, comparing those who could use AI with people who had to work without it.

    In the first experiment, involving 354 people, one group solved basic fractions with ChatGPT available for assistance, while another worked independently, and researchers removed AI access from the first group after 12 questions.

    Although the AI-assisted participants initially achieved greater accuracy, their performance deteriorated sharply when they had to continue without the tool, while those who had worked independently maintained their progress.

    The 2nd experiment

    A second experiment involving 667 participants produced similar results, with people who had relied on AI more likely to answer incorrectly or abandon questions after losing access to assistance.

    The researchers also tested 201 participants on a reading comprehension exercise, finding that AI users again showed reduced accuracy and persistence when they had to work independently.

    The findings suggest that AI can change not only how accurately people complete a task but also how willing they are to keep trying when a problem becomes difficult.

    Christian said the results highlighted the danger of tools that deliver answers instantly, potentially making the slower process of thinking through a problem feel unnecessarily frustrating.

    Researchers describe this effort as “productive struggle”, arguing that working through difficulties can help people develop skills, discover interests and build expertise.

    The implications extend beyond classrooms, as professionals and researchers increasingly use AI to speed up their work, potentially sacrificing opportunities to develop independent judgement and understanding.

    Christian argued that AI systems should be designed to teach users rather than simply provide answers, helping people strengthen their abilities instead of replacing the mental effort needed to acquire them.

    The study, presented at the Conference on Language Modeling in October, does not establish that all AI use damages cognitive abilities, but it raises concerns about what happens when people routinely outsource difficult thinking to machines.

  • AI can transcribe police bodycam footage but struggles to identify who said what

    AI can transcribe police bodycam footage but struggles to identify who said what

    Artificial intelligence (AI) can make hours of police body-camera footage searchable within minutes, but new research warns that transcription errors could distort investigations, undermine accountability and change how encounters between officers and civilians are understood.

    A study published in the ‘Journal of Experimental Criminology’ found that AI transcription software captured much of the language recorded by police body-worn cameras but struggled to identify individual speakers accurately, particularly during complicated exchanges involving several people.

    Researchers examined 176 body-camera videos covering 73 incidents and approximately 23 hours of footage collected by the Kansas City Police Department in Missouri between January and March 2024. They compared transcripts generated by Rev.ai, the speech-recognition technology behind Axon’s automated transcription service, with versions edited by professional human transcribers.

    The AI-generated transcripts identified an average of 3.3 speakers per document, compared with 4.8 in the human-edited versions, and never identified more than seven speakers, while human-edited transcripts recorded as many as 16.

    The software also produced transcripts containing approximately 25,000 fewer words across the sample, suggesting that background conversations, radio communications and contributions from less prominent speakers were frequently omitted.

    What is still wrong

    The problem was not simply that the technology missed words, but that it sometimes attributed statements to the wrong person, potentially changing the meaning of an encounter.

    In one example, a tense exchange between a police officer and a civilian was largely transcribed correctly, but the AI system combined statements from both people and presented them as if they had come from one speaker.

    Such mistakes could have serious consequences when transcripts are used to investigate allegations of misconduct, assess the use of force or evaluate whether officers communicated respectfully with members of the public.

    The researchers argue that AI transcription remains useful for searching large volumes of recordings, identifying keywords and flagging encounters for further examination, but should not replace human review when accurate speaker identification matters.

    The technology offers a practical solution to a significant problem: the Kansas City police department can generate more than 350 hours of body-camera footage in a single day, making comprehensive manual review difficult.

    However, the study has limitations because it examined footage from one police department and assessed a single AI transcription provider, meaning the findings may not apply equally to other systems.

    The researchers also acknowledged that the human-edited transcripts began with AI-generated drafts, rather than being produced independently from the original recordings.

    The findings highlight a broader challenge for artificial intelligence in criminal justice: a system can reproduce most of the words in a recording while still misrepresenting the interaction itself.

  • Google launches AI platform that lets users create video games without coding

    Google launches AI platform that lets users create video games without coding

    Google has launched an experimental platform that allows people to create and play their own video games using artificial intelligence, potentially opening game development to people without coding or technical experience.

    Playground, unveiled on Wednesday, lets users describe a game idea through text prompts and then build it directly in a web browser. Users can start with a blank canvas, modify suggested concepts or use guided prompts to develop their ideas.

    The platform allows creators to change elements including characters, game rules, physics and environments simply by telling the AI what they want to alter. Google said creators remain in control of the creative direction, from the initial concept through to the finished game.

    The games can then be played privately, shared with friends through a link or published to Playground’s community gallery, where other users can discover and play them.

    Google said Playground also supports multiplayer games in selected genres and includes leaderboards, allowing players to compete and follow creators. Games published to the gallery will undergo safety screening based on Google’s community guidelines, with users also able to report content.

    The browser-based platform is designed to work across devices, allowing users to play games on smartphones and laptops rather than requiring specialised software or hardware.

    Google is also planning a more advanced route for creators through Unity Spark, an integration that will bring professional-level game mechanics, high-fidelity 3D capabilities and the Unity runtime to Playground.

    Unity Spark is currently being tested, with a closed beta planned, potentially allowing games that begin as simple AI-generated experiments to evolve into more sophisticated experiences.

    Google described Playground as an early experiment aimed at making game creation more accessible and encouraging everyday creativity.

    The platform launched on Wednesday for users aged 18 and over in the US, with access to game creation being rolled out in tiers based on Google One membership.

  • Letting AI manage your money? Here’s how to avoid losing control

    Letting AI manage your money? Here’s how to avoid losing control

    If an AI agent is managing some or all of your finances, treat it like a highly capable assistant, not a replacement for your judgement. An AI can move money, pay bills, compare investments or make trades, but it can also make mistakes at remarkable speed.

    The first precaution is to limit what the agent can do. Give it only the permissions it needs, much like giving a house sitter one key rather than handing over your entire keyring. An agent that only needs to pay bills should not automatically have permission to transfer large sums or trade investments.

    Set clear financial limits. Put spending caps, transfer limits and approval requirements around important transactions. For example, you might allow an agent to pay a £100 electricity bill automatically but require your approval before it transfers £5,000.

    Keep an emergency brake. Make sure you can quickly revoke the agent’s access, freeze accounts or disable automated transactions. This is similar to having a circuit breaker that cuts electricity when something goes wrong.

    Do not let an AI become the sole guardian of your money. Keep separate access to bank and investment accounts, maintain emergency savings that are not controlled by the agent and review statements regularly. Think of the AI as a co-pilot: it can handle much of the routine work, but you should remain capable of taking control.

    Be particularly careful with investment decisions. AI can analyse huge amounts of information, but it can misunderstand market conditions, rely on inaccurate data or act on instructions it has interpreted incorrectly. Require human approval for large or unusual trades.

    Finally, protect the AI account itself. Use a strong, unique password and multi-factor authentication, and be cautious about connecting new services or approving unfamiliar requests.

    The central rule is simple: automate routine decisions, but keep humans in charge of consequential ones. The more money an AI can move, the more safeguards should stand between its recommendation and your bank account.

    Here’s the quick checklist

    • Limit the AI agent’s access to only the accounts and services it needs.
    • Set spending, transfer and trading limits, with human approval required for large transactions.
    • Keep an emergency brake so you can quickly revoke access or freeze transactions.
    • Maintain separate savings and accounts that the AI cannot control.
    • Review bank statements, payments and investment activity regularly.
    • Require human approval for large, unusual or high-risk investments.
    • Check the information the AI uses before acting on major financial decisions.
    • Protect the AI account with a strong, unique password and multi-factor authentication.
    • Be cautious when connecting new financial services or approving unfamiliar requests.
    • Treat the AI as a co-pilot, not the person in charge of your finances.
  • Reality drift: when AI makes the real world harder to believe

    Reality drift: when AI makes the real world harder to believe

    It’s AI Lingo time, folks.

    Today, we try and understand what “Reality drift” means.

    It describes the gradual erosion of the boundary between what is created by humans and what is generated by artificial intelligence (AI), making it increasingly difficult to distinguish authentic material from synthetic content.

    The term applies to a world in which AI-generated images, videos, voices, articles and social media posts become sufficiently convincing that people can no longer rely on appearance, sound or writing style to establish whether something is real.

    Reality drift is different from a conventional deepfake because it does not necessarily involve deliberate deception, and instead describes a broader change in the information environment.

    As generative AI becomes embedded in search engines, social networks, advertising, entertainment and news production, synthetic material can become part of everyday communication without being clearly identified as machine-generated.

    This creates a problem of accumulated uncertainty, because people may begin questioning genuine material as well as fabricated content.

    A photograph of a political event, for example, can be dismissed as AI-generated even when it is authentic, while a synthetic image can acquire credibility simply because it appears plausible.

    The effect is not only technological but psychological, as repeated exposure to synthetic media can change people’s expectations about what reality looks and sounds like.

    Reality drift can also affect journalism, politics and personal relationships by weakening the evidential value of photographs, recordings and written testimony.

    The central danger is therefore not that AI will make every piece of information false, but that it could make certainty increasingly difficult to achieve.

  • Data centres in space: can orbital computing power the AI boom?

    Data centres in space: can orbital computing power the AI boom?

    The race to build data centres in space is moving from science fiction towards an engineering experiment, driven by the enormous computing power needed to train and run artificial intelligence (AI) models.

    The basic idea is straightforward: put computing hardware on satellites or large orbital platforms, connect them through high-speed communications networks and use the vacuum of space as an environment for cooling and solar energy as the power source.

    Companies developing the concept argue that space could eventually offer advantages that are difficult to replicate on Earth, where data centres consume huge amounts of electricity and water and require increasingly large amounts of land and grid capacity.

    Solar panels can generate electricity continuously for much of an orbital mission, while radiators can release waste heat directly into space, potentially reducing the need for the water-intensive cooling systems used by many terrestrial data centres.

    The biggest obstacle, however, is getting the hardware into orbit and keeping it there.

    Rocket launches remain expensive, although costs have fallen dramatically, while computing equipment must withstand radiation, extreme temperature changes and the physical stresses of launch, potentially requiring specialised hardware that is more expensive and less powerful than its Earth-based equivalent.

    There is also the problem of maintenance. A failed server in a conventional data centre can be replaced within hours, whereas repairing or upgrading equipment hundreds of kilometres above Earth is considerably harder.

    Moving data between space and Earth presents another challenge, particularly for AI applications that require enormous volumes of information to be transferred rapidly.

    For those reasons, space-based computing is unlikely to replace conventional data centres soon, but that does not mean the concept will fail.

    The most plausible early applications are specialised workloads, such as processing satellite imagery in orbit before transmitting results to Earth, rather than running massive consumer AI platforms entirely from space.

    If launch costs continue falling and radiation-resistant chips, autonomous maintenance and optical communications improve, orbital data centres could eventually become economically viable for selected workloads.

    The question is therefore less whether data centres can operate in space, and more whether the savings in energy, cooling and land can eventually outweigh the extraordinary cost and complexity of putting computing infrastructure into orbit.

  • Nvidia’s new system aims to stop AI agents from going rogue

    Nvidia’s new system aims to stop AI agents from going rogue

    Artificial intelligence (AI) is moving beyond chatbots that simply answer questions, with a new generation of “AI agents” being designed to take actions on behalf of people and businesses.

    These agents can be asked to book a flight, analyse financial information, write and run computer code, manage customer enquiries or complete other complicated tasks with relatively little human involvement.

    That creates a new problem: what happens if an AI agent makes a mistake, behaves unexpectedly or starts doing things it was never supposed to do?

    Nvidia is developing technology aimed at addressing this risk by creating a system that can monitor AI agents while they work and intervene when their behaviour becomes dangerous or goes beyond their instructions.

    The concern is sometimes described as an AI agent “going rogue”. This does not necessarily mean a machine suddenly becoming conscious or deciding to attack humans. In most cases, it means an AI system pursuing a task in an unintended way.

    For example, an agent might be told to find the cheapest way to complete a business task. If given too much freedom, it could potentially access information it should not see, make unauthorised changes to computer systems or take actions that create financial or security risks.

    The more powerful agents become, the bigger this problem could become because they can connect to databases, websites, software and other AI systems.

    How is Nvidia solving this?

    Nvidia’s approach is based on adding a layer of security around AI agents. The system can observe what an agent is attempting to do, assess whether the action is allowed and block or interrupt activity that violates predefined rules.

    Think of it as a security guard standing between an AI agent and the computer systems it has been authorised to use.

    This is different from simply checking an AI’s answer for accuracy. A chatbot giving you a wrong answer is one problem, but an AI agent making a wrong decision and then acting on it can have much more serious consequences.

    The technology is therefore aimed at making AI agents more useful without giving them unlimited freedom.

    For ordinary people, this matters because AI agents are likely to become increasingly common in everyday services, from banking and shopping to workplace software and customer support.

    The basic idea is simple: the more AI can do for us, the more important it becomes to have systems that can stop it when it tries to do something it should not.

  • Here’s a lowdown on “Muse” by Meta

    Here’s a lowdown on “Muse” by Meta

    For those within our community who have still not got this piece of news….

    “Muse” is Meta’s new personal artificial intelligence (AI) agent, designed to do more than simply answer questions. Instead of acting like a chatbot that waits for instructions, Muse is intended to take on tasks, make plans and carry out actions on a user’s behalf.

    Meta introduced Muse in September 2026 as part of its push towards what it calls “personal superintelligence”.

    The simplest way to understand Muse is to think of it as a digital assistant that can actually get things done.

    You could ask it to organise a trip, manage a project, send an email, research a subject or help with a purchase.

    Muse can use a web browser, interact with connected services and continue working on longer tasks after you have closed the app. It can then return to you when it needs approval or when something important changes.

    This makes Muse different from conventional AI chatbots. With a chatbot, you might ask for a holiday itinerary and then have to book the flights and hotels yourself. With an AI agent such as Muse, the goal is for the system to handle much more of that process itself, subject to the permissions you give it.

    Muse can also remember information that matters to you. For example, Meta says it could turn a recipe you saved on Instagram into a shopping list, remember dietary requirements and use that information when helping organise a dinner.

    A few days ago, Meta launched “Muse Charm”.

    This is a small, keychain-sized physical device designed to give people instant access to their personal Muse AI agent without having to unlock a phone or open an app.

    Zuckerberg described it as a way to talk to Muse and show it what is happening around you.

    The device has a small display showing Muse’s animated avatar, known internally as Jolly, and a fingerprint sensor that lets users start talking to Muse. Meta says it plans to ship the device in December, although it has not announced a price.

    Here is how Meta has demonstrated it working:

    1. You tap the fingerprint sensor.
      The sensor is built into the corner of the device. Tapping it lets you start talking to Muse without unlocking your phone or opening an app. Zuckerberg demonstrated exactly this at Meta Connect.
    2. You speak naturally to Muse.
      Charm contains Meta’s real-time voice and avatar technology, so the interaction is intended to feel like having a conversation rather than typing questions into a chatbot. The small screen displays a visual representation of your Muse while you talk.
    3. Muse does the actual AI work.
      This is the key distinction. Charm isn’t primarily there to provide computing power. It gives you a quick physical interface for reaching Muse, while Muse can perform tasks using Meta’s connected services. Meta describes Muse as an agent that can continue working on tasks in the background.
    4. Charm can give Muse visual information.
      Zuckerberg said it would be a fast way to “show it what’s going on around you”. That means, for example, you could point the device towards something and ask Muse about it, rather than having to describe what you are seeing. Meta has not yet published all the technical details of how the camera and cloud processing will work.
    5. It can work away from Wi-Fi.
      Reports from the Connect announcement say the device includes cellular connectivity, allowing it to access Muse while you’re out and about.

    So why not just use your phone?

    That’s actually the interesting part.

    With a phone, the sequence is take out phone → unlock it → find the app → start talking.

    With Charm, Meta is trying to reduce that to tap → talk.

    Zuckerberg explicitly described it as the fastest way to reach Muse for people who aren’t wearing Meta’s AI glasses.

    What about privacy?

    Meta says privacy and security are central to how Muse works.

    It runs inside a dedicated virtual computer, with controls designed to prevent it from taking sensitive actions without permission. Users decide which services Muse can access and what it is allowed to do.

    Meta is also bringing Muse to its AI glasses, allowing users to interact with their personal agent hands-free and eventually ask it to act on what they are looking at.

    Image credit: Meta

  • Bring Your Own Agent to Android Studio, says Google

    Bring Your Own Agent to Android Studio, says Google

    Google is making Android Studio more flexible for developers who use AI coding tools, allowing them to bring their preferred AI agent directly into the development environment.

    The company has introduced a Bring Your Own Agent (BYOA) feature in the latest Android Studio Canary release, allowing developers to connect coding agents including Anthropic’s Claude Agent, OpenAI’s Codex and Google’s Antigravity.

    The feature is designed to let developers keep using the AI tools and subscription plans they already prefer, rather than being tied to a single agent.

    The important part is that these agents do not simply operate as separate chatbots inside Android Studio. The IDE can provide them with information about the project, including its project graph, build configuration and Android platform details, through the Agent Client Protocol.

    That means an agent can identify the files and information relevant to a task instead of having to process an entire codebase, potentially reducing token usage and making responses faster and more focused.

    Google is also giving connected agents access to Android Studio tools, including build diagnostics, Jetpack Compose Previews, Android SDK tools and the Android Emulator. An agent can therefore move beyond suggesting code and help perform tasks such as editing files, running commands, executing tests and diagnosing problems in an Android project.

    What this means for developers

    For developers, this could make the IDE more like a central workspace for multiple AI agents. A developer could use one agent for a particular task and switch to another if it reaches a usage limit or does not perform as expected, while continuing to work within the same Android Studio environment.

    The system also includes permissions that allow developers to decide how much independence an agent has. Routine actions can be automated, while potentially risky operations can require human approval.

    Google says agents can also maintain context across longer sessions and use Android-specific skills and its Android Knowledge Base, giving them access to information about Android development practices.

    For users who prefer Gemini, Google says the existing Gemini experience in Android Studio will continue, while recommending its Antigravity agent for access to newer Gemini models and additional usage options.

    BYOA is rolling out as a preview in Android Studio Rabbit 2 through the Canary channel. Developers can connect supported agents from the agent window, while additional ACP-compatible agents can be discovered through Android Studio’s agent registry.

    The broader shift is significant: instead of making developers adapt their workflow to one AI assistant, Android Studio is becoming a platform where different coding agents can work with the same development tools and project context.

  • Why banks are sounding the alarm on your new AI shopping assistants

    Why banks are sounding the alarm on your new AI shopping assistants

    Imagine you have a little digital butler living in your phone. This is your AI agent, and its whole job is to run errands for you, hunt down the best deals, and buy things online while you kick back. Sounds amazing, right?

    Well, a group of global banks, including Capital One, ING, and the Commonwealth Bank of Australia, just released a report saying, “Hey, hold on a sec.” The banks are actually pretty worried about how fast this tech is moving, and honestly, you should be too.

    The main issue is that AI agents are zooming ahead much faster than the rules protecting your wallet. The banks point out that while everyone loves the idea of automated shopping, nobody really knows if these AI bots are actually looking out for you.

    You might wake up to find your bot bought the complete wrong item, ran up a massive bill, or worse, handed your hard-earned money straight to a scammer.

    Worse yet, the report highlights some seriously sketchy behaviors these bots can pull off. Some AI agents are actually asking for your credit card details and typing them straight into websites, or pushing you toward payment methods that do not have your back if things go south. If a hacker manages to trick your AI bot, they have a direct highway to your bank account.

    Click here to read the rest of this newsletter.