Imagine chatting with an AI about a health concern, and instead of just processing your words, it could actually look at the photo of that worrying rash or make sense of your ECG printout. That’s what Google is aiming for.

We already knew AMIE showed promise in text-based medical chats, thanks to earlier work published in Nature. But let’s face it, real medicine isn’t just about words.

Doctors rely heavily on what they can see – skin conditions, readings from machines, lab reports. As the Google team rightly points out, even simple instant messaging platforms “allow static multimodal information (e.g., images and documents) to enrich discussions.”

Text-only AI was missing a huge piece of the puzzle. The big question, as the researchers put it, was “Whether LLMs can conduct diagnostic clinical conversations that incorporate this more complex type of information.”

Google teaches AMIE to look and reason

Google’s engineers have beefed up AMIE using their Gemini 2.0 Flash model as the brains of the operation. They’ve combined this with what they call a “state-aware reasoning framework.” In plain English, this means the AI doesn’t just follow a script; it adapts its conversation based on what it’s learned so far and what it still needs to figure out.

It’s close to how a human clinician works: gathering clues, forming ideas about what might be wrong, and then asking for more specific information – including visual evidence – to narrow things down.

“This enables AMIE to request relevant multimodal artifacts when needed, interpret their findings accurately, integrate this information seamlessly into the ongoing dialogue, and use it to refine diagnoses,” Google explains.

Think of the conversation flowing through stages: first gathering the patient’s history, then moving towards diagnosis and management suggestions, and finally follow-up. The AI constantly assesses its own understanding, asking for that skin photo or lab result if it senses a gap in its knowledge.

To get this right without endless trial-and-error on real people, Google built a detailed simulation lab.

Google created lifelike patient cases, pulling realistic medical images and data from sources like the PTB-XL ECG database and the SCIN dermatology image set, adding plausible backstories using Gemini. Then, they let AMIE ‘chat’ with simulated patients within this setup and automatically check how well it performed on things like diagnostic accuracy and avoiding errors (or ‘hallucinations’).

The virtual OSCE: Google puts AMIE through its paces

The real test came in a setup designed to mirror how medical students are assessed: the Objective Structured Clinical Examination (OSCE).

Google ran a remote study involving 105 different medical scenarios. Real actors, trained to portray patients consistently, interacted either with the new multimodal AMIE or with actual human primary care physicians (PCPs). These chats happened through an interface where the ‘patient’ could upload images, just like you might in a modern messaging app.

Afterwards, specialist doctors (in dermatology, cardiology, and internal medicine) and the patient actors themselves reviewed the conversations.

The human doctors scored everything from how well history was taken, the accuracy of the diagnosis, the quality of the suggested management plan, right down to communication skills and empathy—and, of course, how well the AI interpreted the visual information.

Surprising results from the simulated clinic

Here’s where it gets really interesting. In this head-to-head comparison within the controlled study environment, Google found AMIE didn’t just hold its own—it often came out ahead.

The AI was rated as being better than the human PCPs at interpreting the multimodal data shared during the chats. It also scored higher on diagnostic accuracy, producing differential diagnosis lists (the ranked list of possible conditions) that specialists deemed more accurate and complete based on the case details.

Specialist doctors reviewing the transcripts tended to rate AMIE’s performance higher across most areas. They particularly noted “the quality of image interpretation and reasoning,” the thoroughness of its diagnostic workup, the soundness of its management plans, and its ability to flag when a situation needed urgent attention.

Perhaps one of the most surprising findings came from the patient actors: they often found the AI to be more empathetic and trustworthy than the human doctors in these text-based interactions.

And, on a critical safety note, the study found no statistically significant difference between how often AMIE made errors based on the images (hallucinated findings) compared to the human physicians.

Technology never stands still, so Google also ran some early tests swapping out the Gemini 2.0 Flash model for the newer Gemini 2.5 Flash.

Using their simulation framework, the results hinted at further gains, particularly in getting the diagnosis right (Top-3 Accuracy) and suggesting appropriate management plans.

While promising, the team is quick to add a dose of realism: these are just automated results, and “rigorous assessment through expert physician review is essential to confirm these performance benefits.”

Important reality checks

Google is commendably upfront about the limitations here. “This study explores a research-only system in an OSCE-style evaluation using patient actors, which substantially under-represents the complexity… of real-world care,” they state clearly. 

Simulated scenarios, however well-designed, aren’t the same as dealing with the unique complexities of real patients in a busy clinic. They also stress that the chat interface doesn’t capture the richness of a real video or in-person consultation.

So, what’s the next step? Moving carefully towards the real world. Google is already partnering with Beth Israel Deaconess Medical Center for a research study to see how AMIE performs in actual clinical settings with patient consent.

The researchers also acknowledge the need to eventually move beyond text and static images towards handling real-time video and audio—the kind of interaction common in telehealth today.

Giving AI the ability to ‘see’ and interpret the kind of visual evidence doctors use every day offers a glimpse of how AI might one day assist clinicians and patients. However, the path from these promising findings to a safe and reliable tool for everyday healthcare is still a long one that requires careful navigation.

(Photo by Alexander Sinn)

See also: Are AI chatbots really changing the world of work?

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Google AMIE: AI doctor learns to ‘see’ medical images appeared first on AI News.

Despite huge and ongoing advancements in generative AI, the massive wave it was supposed to create in the world of work looks more like a ripple so far.

Researchers Anders Humlum (University of Chicago) and Emilie Vestergaard (University of Copenhagen) didn’t just rely on anecdotes. They dug deep, connecting responses from two big surveys (late 2023 and 2024) with official, detailed records about jobs and pay in Denmark.

The pair zoomed in on around 25,000 people working in 7,000 different places, covering 11 jobs thought to be right in the path of AI disruption.   

Everyone’s using AI chatbots for work, but where are the benefits?

What they found confirms what many of us see: AI chatbots are everywhere in Danish workplaces now. Most bosses are actually encouraging staff to use them, a real turnaround from the early days when companies were understandably nervous about things like data privacy.

Almost four out of ten employers have even rolled out their own in-house chatbots, and nearly a third of employees have had some formal training on these tools.   

When bosses gave the nod, the number of staff using chatbots practically doubled, jumping from 47% to 83%. It also helped level the playing field a bit. That gap between men and women using chatbots? It shrank noticeably when companies actively encouraged their use, especially when they threw in some training.

So, the tools are popular, companies are investing, people are getting trained… but the big economic shift? It seems to be missing in action.

Using statistical methods to compare people who used AI chatbots for work with those who didn’t, both before and after ChatGPT burst onto the scene, the researchers found… well, basically nothing.

“Precise zeros,” the researchers call their findings. No significant bump in pay, no change in recorded work hours, across all 11 job types they looked at. And they’re pretty confident about this – the numbers rule out any average effect bigger than just 1%.

This wasn’t just a blip, either. The lack of impact held true even for the keen beans who jumped on board early, those using chatbots daily, or folks working where the boss was actively pushing the tech.

Looking at whole workplaces didn’t change the story; places with lots of chatbot users didn’t see different trends in hiring, overall wages, or keeping staff compared to places using them less.

Productivity gains: More of a gentle nudge than a shove

Why the big disconnect? Why all the hype and investment if it’s not showing up in paychecks or job stats? The study flags two main culprits: the productivity boosts aren’t as huge as hoped in the real world, and what little gains there are aren’t really making their way into wages.

Sure, people using AI chatbots for work felt they were helpful. They mentioned better work quality and feeling more creative. But the number one benefit? Saving time.

However, when the researchers crunched the numbers, the average time saved was only about 2.8% of a user’s total work hours. That’s miles away from the huge 15%, 30%, even 50% productivity jumps seen in controlled lab-style experiments (RCTs) involving similar jobs.

Why the difference? A few things seem to be going on. Those experiments often focus on jobs or specific tasks where chatbots really shine (like coding help or basic customer service responses). This study looked at a wider range, including jobs like teaching where the benefits might be smaller.

The researchers stress the importance of what they call “complementary investments”. People whose companies encouraged chatbot use and provided training actually did report bigger benefits – saving more time, improving quality, and feeling more creative. This suggests that just having the tool isn’t enough; you need the right support and company environment to really unlock its potential.

And even those modest time savings weren’t padding wallets. The study reckons only a tiny fraction – maybe 3% to 7% – of the time saved actually showed up as higher earnings. It might be down to standard workplace inertia, or maybe it’s just harder to ask for a raise based on using a tool your boss hasn’t officially blessed, especially when many people started using them off their own bat.

Making new work, not less work

One fascinating twist is that AI chatbots aren’t just about doing old work tasks faster. They seem to be creating new tasks too. Around 17% of people using them said they had new workloads, mostly brand new types of tasks.

This phenomenon happened more often in workplaces that encouraged chatbot use. It even spilled over to people not using the tools – about 5% of non-users reported new tasks popping up because of AI, especially teachers having to adapt assignments or spot AI-written homework.   

What kind of new tasks? Things like figuring out how to weave AI into daily workflows, drafting content with AI help, and importantly, dealing with the ethical side and making sure everything’s above board. It hints that companies are still very much in the ‘figuring it out’ phase, spending time and effort adapting rather than just reaping instant rewards.

What’s the verdict on the work impact of AI chatbots?

The researchers are careful not to write off generative AI completely. They see pathways for it to become more influential over time, especially as companies get better at integrating it and maybe as those “new tasks” evolve.

But for now, their message is clear: the current reality doesn’t match the hype about a massive, immediate job market overhaul.

“Despite rapid adoption and substantial investments… our key finding is that AI chatbots have had minimal impact on productivity and labor market outcomes to date,” the researchers conclude.   

It brings to mind that old quote about the early computer age: seen everywhere, except in the productivity stats. Two years on from ChatGPT’s launch kicking off the fastest tech adoption we’ve ever seen, its actual mark on jobs and pay looks surprisingly light.

The revolution might still be coming, but it seems to be taking its time.   

See also: Claude Integrations: Anthropic adds AI to your favourite work tools

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Are AI chatbots really changing the world of work? appeared first on AI News.

Starting with Integrations, the feature builds on a technical standard Anthropic released last year (the Model Context Protocol, or MCP), but makes it much easier to use. Before, setting this up was a bit technical and local. Now, developers can build secure bridges allowing Claude to connect safely with apps over the web or on your desktop.

For end-users of Claude, this means you can now hook it up to a growing list of popular work software. Right out of the gate, they’ve included support for ten big names: Atlassian’s Jira and Confluence (hello, project managers and dev teams!), the automation powerhouse Zapier, Cloudflare, customer comms tool Intercom, plus Asana, Square, Sentry, PayPal, Linear, and Plaid. Stripe and GitLab are joining the party soon.

So, what’s the big deal? The real advantage here is context. When Claude can see your project history in Jira, read your team’s knowledge base in Confluence, or check task updates in Asana, it stops guessing and starts understanding what you’re working on.

“When you connect your tools to Claude, it gains deep context about your work—understanding project histories, task statuses, and organisational knowledge—and can take actions across every surface,” explains Anthropic.

They add, “Claude becomes a more informed collaborator, helping you execute complex projects in one place with expert assistance at every step.”

Let’s look at what this means in practice. Connect Zapier, and you suddenly give Claude the keys to thousands of apps linked by Zapier’s workflows. You could just ask Claude, conversationally, to trigger a complex sequence – maybe grab the latest sales numbers from HubSpot, check your calendar, and whip up some meeting notes, all without you lifting a finger in those apps.

For teams using Atlassian’s Jira and Confluence, Claude could become a serious helper. Think drafting product specs, summarising long Confluence documents so you don’t have to wade through them, or even creating batches of linked Jira tickets at once. It might even spot potential roadblocks by analysing project data.

And if you use Intercom for customer chats, this integration could be a game-changer. Intercom’s own AI assistant, Fin, can now work with Claude to do things like automatically create a bug report in Linear if a customer flags an issue. You could also ask Claude to sift through your Intercom chat history to spot patterns, help debug tricky problems, or summarise what customers are saying – making the whole journey from feedback to fix much smoother.

Anthropic is also making it easier for developers to build even more of these connections. They reckon that using their tools (or platforms like Cloudflare that handle the tricky bits like security and setup), developers can whip up a custom Integration with Claude in about half an hour. This could mean connecting Claude to your company’s unique internal systems or specialised industry software.

Beyond tool integrations, Claude gets a serious research upgrade

Alongside these new connections, Anthropic has given Claude’s Research feature a serious boost. It could already search the web and your Google Workspace files, but the new ‘Advanced Research’ mode is built for when you need to dig really deep.

Flip the switch for this advanced mode, and Claude tackles big questions differently. Instead of just one big search, it intelligently breaks your request down into smaller chunks, investigates each part thoroughly – using the web, your Google Docs, and now tapping into any apps you’ve connected via Integrations – before pulling it all together into a detailed report.

Now, this deeper digging takes a bit more time. While many reports might only take five to fifteen minutes, Anthropic says the really complex investigations could have Claude working away for up to 45 minutes. That might sound like a while, but compare it to the hours you might spend grinding through that research manually, and it starts to look pretty appealing.

Importantly, you can trust the results. When Claude uses information from any source – whether it’s a website, an internal doc, a Jira ticket, or a Confluence page – it gives you clear links straight back to the original. No more wondering where the AI got its information from; you can check it yourself.

These shiny new Integrations and the Advanced Research mode are rolling out now in beta for folks on Anthropic’s paid Max, Team, and Enterprise plans. If you’re on the Pro plan, don’t worry – access is coming your way soon.

Also worth noting: the standard web search feature inside Claude is now available everywhere, for everyone on any paid Claude.ai plan (Pro and up). No more geographical restrictions on that front.

Putting it all together, these updates and integrations show Anthropic is serious about making Claude genuinely useful in a professional context. By letting it plug directly into the tools we already use and giving it more powerful ways to analyse information, they’re pushing Claude towards being less of a novelty and more of an essential part of the modern toolkit.

(Image credit: Anthropic)

See also: Baidu ERNIE X1 and 4.5 Turbo boast high performance at low cost

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Claude Integrations: Anthropic adds AI to your favourite work tools appeared first on AI News.

The improved security tools for the Llama AI models arrive alongside fresh resources from Meta designed to help cybersecurity teams harness AI for defence. It’s all part of their push to make developing and using AI a bit safer for everyone involved.

Developers working with the Llama family of models now have some upgraded kit to play with. You can grab these latest Llama Protection tools directly from Meta’s own Llama Protections page, or find them where many developers live: Hugging Face and GitHub.

First up is Llama Guard 4. Think of it as an evolution of Meta’s customisable safety filter for AI. The big news here is that it’s now multimodal so it can understand and apply safety rules not just to text, but to images as well. That’s crucial as AI applications get more visual. This new version is also being baked into Meta’s brand-new Llama API, which is currently in a limited preview.

Then there’s LlamaFirewall. This is a new piece of the puzzle from Meta, designed to act like a security control centre for AI systems. It helps manage different safety models working together and hooks into Meta’s other protection tools. Its job? To spot and block the kind of risks that keep AI developers up at night – things like clever ‘prompt injection’ attacks designed to trick the AI, potentially dodgy code generation, or risky behaviour from AI plug-ins.

Meta has also given its Llama Prompt Guard a tune-up. The main Prompt Guard 2 (86M) model is now better at sniffing out those pesky jailbreak attempts and prompt injections. More interestingly, perhaps, is the introduction of Prompt Guard 2 22M.

Prompt Guard 2 22M is a much smaller, nippier version. Meta reckons it can slash latency and compute costs by up to 75% compared to the bigger model, without sacrificing too much detection power. For anyone needing faster responses or working on tighter budgets, that’s a welcome addition.

But Meta isn’t just focusing on the AI builders; they’re also looking at the cyber defenders on the front lines of digital security. They’ve heard the calls for better AI-powered tools to help in the fight against cyberattacks, and they’re sharing some updates aimed at just that.

The CyberSec Eval 4 benchmark suite has been updated. This open-source toolkit helps organisations figure out how good AI systems actually are at security tasks. This latest version includes two new tools:

• CyberSOC Eval: Built with the help of cybersecurity experts CrowdStrike, this framework specifically measures how well AI performs in a real Security Operation Centre (SOC) environment. It’s designed to give a clearer picture of AI’s effectiveness in threat detection and response. The benchmark itself is coming soon.
• AutoPatchBench: This benchmark tests how good Llama and other AIs are at automatically finding and fixing security holes in code before the bad guys can exploit them.

To help get these kinds of tools into the hands of those who need them, Meta is kicking off the Llama Defenders Program. This seems to be about giving partner companies and developers special access to a mix of AI solutions – some open-source, some early-access, some perhaps proprietary – all geared towards different security challenges.

As part of this, Meta is sharing an AI security tool they use internally: the Automated Sensitive Doc Classification Tool. It automatically slaps security labels on documents inside an organisation. Why? To stop sensitive info from walking out the door, or to prevent it from being accidentally fed into an AI system (like in RAG setups) where it could be leaked.

They’re also tackling the problem of fake audio generated by AI, which is increasingly used in scams. The Llama Generated Audio Detector and Llama Audio Watermark Detector are being shared with partners to help them spot AI-generated voices in potential phishing calls or fraud attempts. Companies like ZenDesk, Bell Canada, and AT&T are already lined up to integrate these.

Finally, Meta gave a sneak peek at something potentially huge for user privacy: Private Processing. This is new tech they’re working on for WhatsApp. The idea is to let AI do helpful things like summarise your unread messages or help you draft replies, but without Meta or WhatsApp being able to read the content of those messages.

Meta is being quite open about the security side, even publishing their threat model and inviting security researchers to poke holes in the architecture before it ever goes live. It’s a sign they know they need to get the privacy aspect right.

Overall, it’s a broad set of AI security announcements from Meta. They’re clearly trying to put serious muscle behind securing the AI they build, while also giving the wider tech community better tools to build safely and defend effectively.

See also: Alarming rise in AI-powered scams: Microsoft reveals $4B in thwarted fraud

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Meta beefs up AI security with new Llama tools  appeared first on AI News.

So, what’s the big deal about E-Beam lithography? Imagine trying to draw incredibly complex patterns, but thousands of times smaller than a human hair. That’s essentially what this technology does, using a focused beam of tiny electrons.

Such precision is vital for designing the microscopic components inside the chips that run everything from our smartphones and gaming consoles to life-saving medical scanners and advanced defence systems.

Semiconductors are already big business for the UK, adding around £10 billion to its economy each year. And that figure is only expected to climb, potentially hitting £17 billion by the end of the decade.

Nurturing this sector is a major opportunity for the UK—not just for bragging rights in advanced manufacturing, but for creating high-value jobs and driving real economic growth.

Speaking at the launch of the facility in Southampton, Science Minister Lord Patrick Vallance said: “Britain is home to some of the most exciting semiconductor research anywhere in the world—and Southampton’s new E-Beam facility is a major boost to our national capabilities.

“By investing in both infrastructure and talent, we’re giving our researchers and innovators the support they need to develop next-generation chips right here in the UK.”

Lord Vallance’s visit wasn’t just a photo opportunity, though. It came alongside some sobering news: fresh research published today highlights that one of the biggest hurdles facing the UK’s growing chip industry is finding enough people with the right skills.

We’re talking about a serious talent crunch. When you consider that a single person working in semiconductors contributes an average of £460,000 to the economy each year, you can see why plugging this skills gap is so critical.

So, what’s the plan? The government isn’t just acknowledging the problem; they’re putting money where their mouth is with a £4.75 million semiconductor skills package. The idea is to build up that talent pipeline, making sure universities like Southampton – already powerhouses of chip innovation – have resources like the E-Beam lab and the students they need.

“Our £4.75 million skills package will support our Plan for Change by helping more young people into high-value semiconductors careers, closing skills gaps and backing growth in this critical sector,” Lord Vallance explained.

Here’s where that cash is going:

• Getting students hooked (£3 million): Fancy £5,000 towards your degree? 300 students starting Electronics and Electrical Engineering courses this year will get just that, along with specific learning modules to show them what a career in semiconductors actually involves, particularly in chip design and making the things.
• Practical chip skills (£1.2 million): It’s one thing learning the theory, another designing a real chip. This pot will fund new hands-on chip design courses for students (undergrad and postgrad) and even train up lecturers. They’re also looking into creating conversion courses to tempt talented people from other fields into the chip world.
• Inspiring the next generation (Nearly £550,000): To really build a long-term pipeline, you need to capture interest early. This funding aims to give 7,000 teenagers (15-18) and 450 teachers some real, hands-on experience with semiconductors, working with local companies in existing UK chip hotspots like Newport, Cambridge, and Glasgow. The goal is to show young people the cool career paths available right on their doorstep.

Ultimately, the hope is that this targeted support will give the UK semiconductor scene the skilled workforce it needs to thrive. It’s about encouraging more students to jump into these valuable careers, helping companies find the people they desperately need, and making sure the UK stays at the forefront of the technologies that will shape tomorrow’s economy.

Professor Graham Reed, who heads up the Optoelectronics Research Centre (ORC) at Southampton University, commented: “The introduction of the new E-Beam facility will reinforce our position of hosting the most advanced cleanroom in UK academia.

“It facilitates a vast array of innovative and industrially relevant research, and much needed semiconductor skills training.”

Putting world-class tools in the hands of researchers while simultaneously investing in the people who will use them will help to cement the UK’s leadership in semiconductors.

See also: AI in education: Balancing promises and pitfalls

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post UK opens Europe’s first E-Beam semiconductor chip lab appeared first on AI News.

There’s a real push to bring AI into schools, and you can see why. The recent executive order on youth education from President Trump recognised that if future generations are going to do well in an increasingly automated world, they need to be ready.

“To ensure the United States remains a global leader in this technological revolution, we must provide our nation’s youth with opportunities to cultivate the skills and understanding necessary to use and create the next generation of AI technology,” President Trump declared.

So, what does AI actually look like in the classroom?

One of the biggest hopes for AI in education is making learning more personal. Imagine software that can figure out how individual students are doing, then adjust the pace and materials just for them. This could mean finally moving away from the old one-size-fits-all approach towards learning environments that adapt and offer help exactly where it’s needed.

The US executive order hints at this, wanting to improve results through things like “AI-based high-quality instructional resources” and “high-impact tutoring.”

And what about teachers? AI could be a huge help here too, potentially taking over tedious admin tasks like grading, freeing them up to actually teach. Plus, AI software might offer fresh ways to present information.

Getting kids familiar with AI early on could also take away some of the mystery around the technology. It might spark their “curiosity and creativity” and give them the foundation they need to become “active and responsible participants in the workforce of the future.”

The focus stretches to lifelong learning and getting people ready for the job market. On top of that, AI tools like text-to-speech or translation features can make learning much more accessible for students with disabilities, opening up educational environments for everyone.

Not all smooth sailing: The challenges ahead for AI in education

While the potential is huge, we need to be realistic about the significant hurdles and potential downsides.

First off, AI runs on student data – lots of it. That means we absolutely need strong rules and security to make sure this data is collected ethically, used correctly, and kept safe from breaches. Privacy is paramount here.

Then there’s the bias problem. If the data used to train AI reflects existing unfairness in society (and let’s be honest, it often does), the AI could end up repeating or even worsening those inequalities. Think biased assessments or unfair resource allocation. Careful testing and constant checks are crucial to catch and fix this.

We also can’t ignore the digital divide. If some students don’t have reliable internet, the right devices, or the necessary tech infrastructure at home or school, AI could widen the gap between the haves and have-nots. It’s vital that everyone gets fair access.

There’s also a risk that leaning too heavily on AI education tools might stop students from developing essential skills like critical thinking. We need to teach them how to use AI as a helpful tool, not a crutch they can’t function without.

Maybe the biggest piece of the puzzle, though, is making sure our teachers are ready. As the executive order rightly points out, “We must also invest in our educators and equip them with the tools and knowledge.”

This isn’t just about knowing which buttons to push; teachers need to understand how AI fits into teaching effectively and ethically. That requires solid professional development and ongoing support.

A recent GMB Union poll found that while about a fifth of UK schools are using AI now, the staff often aren’t getting the training they need:

View on Threads

Finding the right path forward

It’s going to take everyone – governments, schools, tech companies, and teachers – pulling together in order to ensure that AI plays a positive role in education.

We absolutely need clear policies and standards covering ethics, privacy, bias, and making sure AI is accessible to all students. We also need to keep investing in research to figure out the best ways to use AI in education and to build tools that are fair and effective.

And critically, we need a long-term commitment to teacher education to get educators comfortable and skilled with these changes. Part of this is building broad AI literacy, making sure all students get a basic understanding of this technology and how it impacts society.

AI could be a positive force in education – making it more personalised, efficient, and focused on the skills students actually need. But turning that potential into reality means carefully navigating those tricky ethical, practical, and teaching challenges head-on.

See also: How does AI judge? Anthropic studies the values of Claude

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post AI in education: Balancing promises and pitfalls appeared first on AI News.

Developed as enhancements to the existing ERNIE X1 and 4.5 models, both new Turbo versions highlight multimodal processing, robust reasoning skills, and aggressive pricing strategies designed to capture developer interest and marketshare.

Baidu ERNIE X1 Turbo: Deep reasoning meets cost efficiency

Positioned as a deep-thinking reasoning model, ERNIE X1 Turbo tackles complex tasks requiring sophisticated understanding. It enters a competitive field, claiming superior performance in some benchmarks against rivals like DeepSeek R1, V3, and OpenAI o1:

Key to X1 Turbo’s enhanced capabilities is an advanced “chain of thought” process, enabling more structured and logical problem-solving.

Furthermore, ERNIE X1 Turbo boasts improved multimodal functions – the ability to understand and process information beyond just text, potentially including images or other data types – alongside refined tool utilisation abilities. This makes it particularly well-suited for nuanced applications such as literary creation, complex logical reasoning challenges, code generation, and intricate instruction following.

ERNIE X1 Turbo achieves this performance while undercutting competitor pricing. Input token costs start at $0.14 per million tokens, with output tokens priced at $0.55 per million. This pricing structure is approximately 25% of DeepSeek R1.

Baidu ERNIE 4.5 Turbo: Multimodal muscle at a fraction of the cost

Sharing the spotlight is ERNIE 4.5 Turbo, which focuses on delivering upgraded multimodal features and significantly faster response times compared to its non-Turbo counterpart. The emphasis here is on providing a versatile, responsive AI experience while slashing operational costs.

The model achieves an 80% price reduction compared to the original ERNIE 4.5 with input set at $0.11 per million tokens and output at $0.44 per million tokens. This represents roughly 40% of the cost of the latest version of DeepSeek V3, again highlighting a deliberate strategy to attract users through cost-effectiveness.

Performance benchmarks further bolster its credentials. In multiple tests evaluating both multimodal and text capabilities, Baidu ERNIE 4.5 Turbo outperforms OpenAI’s highly-regarded GPT-4o model. 

In multimodal capability assessments, ERNIE 4.5 Turbo achieved an average score of 77.68 to surpass GPT-4o’s score of 72.76 in the same tests.

While benchmark results always require careful interpretation, this suggests ERNIE 4.5 Turbo is a serious contender for tasks involving an integrated understanding of different data types.

Baidu continues to shake up the AI marketplace

The launch of ERNIE X1 Turbo and 4.5 Turbo signifies a growing trend in the AI sector: the democratisation of high-end capabilities. While foundational models continue to push the boundaries of performance, there is increasing demand for models that balance power with accessibility and affordability.

By lowering the price points for models with sophisticated reasoning and multimodal features, the Baidu ERNIE Turbo series could enable a wider range of developers and businesses to integrate advanced AI into their applications.

This competitive pricing puts pressure on established players like OpenAI and Anthropic, as well as emerging competitors like DeepSeek, potentially leading to further price adjustments across the market.

(Image Credit: Alpha Photo under CC BY-NC 2.0 license)

See also: China’s MCP adoption: AI assistants that actually do things

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Baidu ERNIE X1 and 4.5 Turbo boast high performance at low cost appeared first on AI News.

Training these AI agents presents significant hurdles, particularly when decisions span multiple steps and involve unpredictable feedback from the environment. While reinforcement learning (RL) has shown promise in static tasks like solving maths problems or generating code, its application to dynamic, multi-turn agent training has been less explored.   

Addressing this gap, a collaborative team from institutions including Northwestern University, Stanford University, Microsoft, and New York University has proposed StarPO (State-Thinking-Actions-Reward Policy Optimisation).

StarPO offers a generalised approach for training agents at the trajectory level (i.e. it optimises the entire sequence of interactions, not just individual actions.)

Accompanying this is RAGEN, a modular system built to implement StarPO. This enables the training and evaluation of LLM agents, particularly focusing on their reasoning capabilities under RL. RAGEN provides the necessary infrastructure for rollouts, reward assignment, and optimisation within multi-turn, stochastic (randomly determined) environments.

Minimalist environments, maximum insight

To isolate the core learning challenges from confounding factors like extensive pre-existing knowledge or task-specific engineering, the researchers tested LLMs using RAGEN in three deliberately minimalistic, controllable symbolic gaming environments:   

1. Bandit: A single-turn, stochastic task testing risk-sensitive symbolic reasoning. The agent chooses between options (like ‘Phoenix’ or ‘Dragon’ arms) with different, initially unknown, reward profiles.
2. Sokoban: A multi-turn, deterministic puzzle requiring foresight and planning, as actions (pushing boxes) are irreversible.
3. Frozen Lake: A multi-turn, stochastic grid navigation task where movement attempts can randomly fail, demanding planning under uncertainty.

These environments allow for clear analysis of how agents learn decision-making policies purely through interaction.   

Key findings: Stability, rollouts, and reasoning

The study yielded three significant findings concerning the training of self-evolving LLM agents:

The ‘Echo Trap’ and the need for stability

A recurring problem observed during multi-turn RL training was dubbed the “Echo Trap”. Agents would initially improve but then suffer performance collapse, overfitting to locally rewarded reasoning patterns. 

This was marked by collapsing reward variance, falling entropy (a measure of randomness/exploration), and sudden spikes in gradients (indicating training instability). Early signs included drops in reward standard deviation and output entropy.   

To combat this, the team developed StarPO-S, a stabilised version of the framework. StarPO-S incorporates:   

• Variance-based trajectory filtering: Focusing training on task instances where the agent’s behaviour shows higher uncertainty (higher reward variance), discarding low-variance, less informative rollouts. This improved stability and efficiency.   
• Critic incorporation: Using methods like PPO (Proximal Policy Optimisation), which employ a ‘critic’ to estimate value, generally showed better stability than critic-free methods like GRPO (Group Relative Policy Optimisation) in most tests.   
• Decoupled clipping and KL removal: Techniques adapted from other research (DAPO) involving asymmetric clipping (allowing more aggressive learning from positive rewards) and removing KL divergence penalties (encouraging exploration) further boosted stability and performance.   

StarPO-S consistently delayed collapse and improved final task performance compared to vanilla StarPO.   

Rollout quality is crucial

The characteristics of the ‘rollouts’ (simulated interaction trajectories used for training) significantly impact learning. Key factors identified include:   

• Task diversity: Training with a diverse set of initial states (prompts), but with multiple responses generated per prompt, aids generalisation. The sweet spot seemed to be moderate diversity enabling contrast between different outcomes in similar scenarios.   
• Interaction granularity: Allowing multiple actions per turn (around 5-6 proved optimal) enables better planning within a fixed turn limit, without introducing the noise associated with excessively long action sequences.   
• Rollout frequency: Using fresh, up-to-date rollouts that reflect the agent’s current policy is vital. More frequent sampling (approaching an ‘online’ setting) leads to faster convergence and better generalisation by reducing policy-data mismatch.

Maintaining freshness, alongside appropriate action budgets and task diversity, is key for stable training.   

Reasoning requires careful reward design

Simply prompting models to ‘think’ doesn’t guarantee meaningful reasoning emerges, especially in multi-turn tasks. The study found:

• Reasoning traces helped generalisation in the simpler, single-turn Bandit task, even when symbolic cues conflicted with rewards.   
• In multi-turn tasks like Sokoban, reasoning benefits were limited, and the length of ‘thinking’ segments consistently declined during training. Agents often regressed to direct action selection or produced “hallucinated reasoning” if rewards only tracked task success, revealing a “mismatch between thoughts and environment states.”

This suggests that standard trajectory-level rewards (often sparse and outcome-based) are insufficient. 

“Without fine-grained, reasoning-aware reward signals, agent reasoning hardly emerge[s] through multi-turn RL.”

The researchers propose that future work should explore rewards that explicitly evaluate the quality of intermediate reasoning steps, perhaps using format-based penalties or rewarding explanation quality, rather than just final outcomes.   

RAGEN and StarPO: A step towards self-evolving AI

The RAGEN system and StarPO framework represent a step towards training LLM agents that can reason and adapt through interaction in complex, unpredictable environments.

This research highlights the unique stability challenges posed by multi-turn RL and offers concrete strategies – like StarPO-S’s filtering and stabilisation techniques – to mitigate them. It also underscores the critical role of rollout generation strategies and the need for more sophisticated reward mechanisms to cultivate genuine reasoning, rather than superficial strategies or hallucinations.

While acknowledging limitations – including the need to test on larger models and optimise for domains without easily verifiable rewards – the work opens “a scalable and principled path for building AI systems” in areas demanding complex interaction and verifiable outcomes, such as theorem proving, software engineering, and scientific discovery.

(Image by Gerd Altmann)

See also: How does AI judge? Anthropic studies the values of Claude

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post RAGEN: AI framework tackles LLM agent instability appeared first on AI News.

In an open letter addressed to the Attorneys General of California and Delaware, the group – which also includes legal experts, corporate governance specialists, AI researchers, and nonprofit representatives – argues that the proposed changes fundamentally threaten OpenAI’s original charitable mission.   

OpenAI was founded with a unique structure. Its core purpose, enshrined in its Articles of Incorporation, is “to ensure that artificial general intelligence benefits all of humanity” rather than serving “the private gain of any person.”

The letter’s signatories contend that the planned restructuring – transforming the current for-profit subsidiary (OpenAI-profit) controlled by the original nonprofit entity (OpenAI-nonprofit) into a Delaware public benefit corporation (PBC) – would dismantle crucial governance safeguards.

This shift, the signatories argue, would transfer ultimate control over the development and deployment of potentially transformative Artificial General Intelligence (AGI) from a charity focused on humanity’s benefit to a for-profit enterprise accountable to shareholders.

Original vision of OpenAI: Nonprofit control as a bulwark

OpenAI defines AGI as “highly autonomous systems that outperform humans at most economically valuable work”. While acknowledging AGI’s potential to “elevate humanity,” OpenAI’s leadership has also warned of “serious risk of misuse, drastic accidents, and societal disruption.”

Co-founder Sam Altman and others have even signed statements equating mitigating AGI extinction risks with preventing pandemics and nuclear war.   

The company’s founders – including Altman, Elon Musk, and Greg Brockman – were initially concerned about AGI being developed by purely commercial entities like Google. They established OpenAI as a nonprofit specifically “unconstrained by a need to generate financial return”. As Altman stated in 2017, “The only people we want to be accountable to is humanity as a whole.”

Even when OpenAI introduced a “capped-profit” subsidiary in 2019 to attract necessary investment, it emphasised that the nonprofit parent would retain control and that the mission remained paramount. Key safeguards included:   

• Nonprofit control: The for-profit subsidiary was explicitly “controlled by OpenAI Nonprofit’s board”.   
• Capped profits: Investor returns were capped, with excess value flowing back to the nonprofit for humanity’s benefit.   
• Independent board: A majority of nonprofit board members were required to be independent, holding no financial stake in the subsidiary.   
• Fiduciary duty: The board’s legal duty was solely to the nonprofit’s mission, not to maximising investor profit.   
• AGI ownership: AGI technologies were explicitly reserved for the nonprofit to govern.

Altman himself testified to Congress in 2023 that this “unusual structure” “ensures it remains focused on [its] long-term mission.”

A threat to the mission?

The critics argue the move to a PBC structure would jeopardise these safeguards:   

• Subordination of mission: A PBC board – while able to consider public benefit – would also have duties to shareholders, potentially balancing profit against the mission rather than prioritising the mission above all else.   
• Loss of enforceable duty: The current structure gives Attorneys General the power to enforce the nonprofit’s duty to the public. Under a PBC, this direct public accountability – enforceable by regulators – would likely vanish, leaving shareholder derivative suits as the primary enforcement mechanism.   
• Uncapped profits?: Reports suggest the profit cap might be removed, potentially reallocating vast future wealth from the public benefit mission to private shareholders.   
• Board independence uncertain: Commitments to a majority-independent board overseeing AI development could disappear.   
• AGI control shifts: Ownership and control of AGI would likely default to the PBC and its investors, not the mission-focused nonprofit. Reports even suggest OpenAI and Microsoft have discussed removing contractual restrictions on Microsoft’s access to future AGI.   
• Charter commitments at risk: Commitments like the “stop-and-assist” clause (pausing competition to help a safer, aligned AGI project) might not be honoured by a profit-driven entity.  

OpenAI has publicly cited competitive pressures (i.e. attracting investment and talent against rivals with conventional equity structures) as reasons for the change.

However, the letter counters that competitive advantage isn’t the charitable purpose of OpenAI and that its unique nonprofit structure was designed to impose certain competitive costs in favour of safety and public benefit. 

“Obtaining a competitive advantage by abandoning the very governance safeguards designed to ensure OpenAI remains true to its mission is unlikely to, on balance, advance the mission,” the letter states.   

The authors also question why OpenAI abandoning nonprofit control is necessary merely to simplify the capital structure, suggesting the core issue is the subordination of investor interests to the mission. They argue that while the nonprofit board can consider investor interests if it serves the mission, the restructuring appears aimed at allowing these interests to prevail at the expense of the mission.

Many of these arguments have also been pushed by Elon Musk in his legal action against OpenAI. Earlier this month, OpenAI counter-sued Musk for allegedly orchestrating a “relentless” and “malicious” campaign designed to “take down OpenAI” after he left the company years ago and started rival AI firm xAI.

Call for intervention

The signatories of the open letter urge intervention, demanding answers from OpenAI about how the restructuring away from a nonprofit serves its mission and why safeguards previously deemed essential are now obstacles.

Furthemore, the signatories request a halt to the restructuring, preservation of nonprofit control and other safeguards, and measures to ensure the board’s independence and ability to oversee management effectively in line with the charitable purpose.   

“The proposed restructuring would eliminate essential safeguards, effectively handing control of, and profits from, what could be the most powerful technology ever created to a for-profit entity with legal duties to prioritise shareholder returns,” the signatories conclude.

See also: How does AI judge? Anthropic studies the values of Claude

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Coalition opposes OpenAI shift from nonprofit roots appeared first on AI News.

In a research paper, the Societal Impacts team at Anthropic details a privacy-preserving methodology designed to observe and categorise the values Claude exhibits “in the wild.” This offers a glimpse into how AI alignment efforts translate into real-world behaviour.

The core challenge lies in the nature of modern AI. These aren’t simple programs following rigid rules; their decision-making processes are often opaque.

Anthropic says it explicitly aims to instil certain principles in Claude, striving to make it “helpful, honest, and harmless.” This is achieved through techniques like Constitutional AI and character training, where preferred behaviours are defined and reinforced.

However, the company acknowledges the uncertainty. “As with any aspect of AI training, we can’t be certain that the model will stick to our preferred values,” the research states.

“What we need is a way of rigorously observing the values of an AI model as it responds to users ‘in the wild’ […] How rigidly does it stick to the values? How much are the values it expresses influenced by the particular context of the conversation? Did all our training actually work?”

Analysing Anthropic Claude to observe AI values at scale

To answer these questions, Anthropic developed a sophisticated system that analyses anonymised user conversations. This system removes personally identifiable information before using language models to summarise interactions and extract the values being expressed by Claude. The process allows researchers to build a high-level taxonomy of these values without compromising user privacy.

The study analysed a substantial dataset: 700,000 anonymised conversations from Claude.ai Free and Pro users over one week in February 2025, predominantly involving the Claude 3.5 Sonnet model. After filtering out purely factual or non-value-laden exchanges, 308,210 conversations (approximately 44% of the total) remained for in-depth value analysis.

The analysis revealed a hierarchical structure of values expressed by Claude. Five high-level categories emerged, ordered by prevalence:

1. Practical values: Emphasising efficiency, usefulness, and goal achievement.
2. Epistemic values: Relating to knowledge, truth, accuracy, and intellectual honesty.
3. Social values: Concerning interpersonal interactions, community, fairness, and collaboration.
4. Protective values: Focusing on safety, security, well-being, and harm avoidance.
5. Personal values: Centred on individual growth, autonomy, authenticity, and self-reflection.

These top-level categories branched into more specific subcategories like “professional and technical excellence” or “critical thinking.” At the most granular level, frequently observed values included “professionalism,” “clarity,” and “transparency” – fitting for an AI assistant.

Critically, the research suggests Anthropic’s alignment efforts are broadly successful. The expressed values often map well onto the “helpful, honest, and harmless” objectives. For instance, “user enablement” aligns with helpfulness, “epistemic humility” with honesty, and values like “patient wellbeing” (when relevant) with harmlessness.

Nuance, context, and cautionary signs

However, the picture isn’t uniformly positive. The analysis identified rare instances where Claude expressed values starkly opposed to its training, such as “dominance” and “amorality.”

Anthropic suggests a likely cause: “The most likely explanation is that the conversations that were included in these clusters were from jailbreaks, where users have used special techniques to bypass the usual guardrails that govern the model’s behavior.”

Far from being solely a concern, this finding highlights a potential benefit: the value-observation method could serve as an early warning system for detecting attempts to misuse the AI.

The study also confirmed that, much like humans, Claude adapts its value expression based on the situation.

When users sought advice on romantic relationships, values like “healthy boundaries” and “mutual respect” were disproportionately emphasised. When asked to analyse controversial history, “historical accuracy” came strongly to the fore. This demonstrates a level of contextual sophistication beyond what static, pre-deployment tests might reveal.

Furthermore, Claude’s interaction with user-expressed values proved multifaceted:

• Mirroring/strong support (28.2%): Claude often reflects or strongly endorses the values presented by the user (e.g., mirroring “authenticity”). While potentially fostering empathy, the researchers caution it could sometimes verge on sycophancy.
• Reframing (6.6%): In some cases, especially when providing psychological or interpersonal advice, Claude acknowledges the user’s values but introduces alternative perspectives.
• Strong resistance (3.0%): Occasionally, Claude actively resists user values. This typically occurs when users request unethical content or express harmful viewpoints (like moral nihilism). Anthropic posits these moments of resistance might reveal Claude’s “deepest, most immovable values,” akin to a person taking a stand under pressure.

Limitations and future directions

Anthropic is candid about the method’s limitations. Defining and categorising “values” is inherently complex and potentially subjective. Using Claude itself to power the categorisation might introduce bias towards its own operational principles.

This method is designed for monitoring AI behaviour post-deployment, requiring substantial real-world data and cannot replace pre-deployment evaluations. However, this is also a strength, enabling the detection of issues – including sophisticated jailbreaks – that only manifest during live interactions.

The research concludes that understanding the values AI models express is fundamental to the goal of AI alignment.

“AI models will inevitably have to make value judgments,” the paper states. “If we want those judgments to be congruent with our own values […] then we need to have ways of testing which values a model expresses in the real world.”

This work provides a powerful, data-driven approach to achieving that understanding. Anthropic has also released an open dataset derived from the study, allowing other researchers to further explore AI values in practice. This transparency marks a vital step in collectively navigating the ethical landscape of sophisticated AI.

We’ve made the dataset of Claude’s expressed values open for anyone to download and explore for themselves.

Download the data: https://t.co/rxwPsq6hXf

— Anthropic (@AnthropicAI) April 21, 2025

See also: Google introduces AI reasoning control in Gemini 2.5 Flash

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post How does AI judge? Anthropic studies the values of Claude appeared first on AI News.

The latest releases from Meta focus heavily on enhancing AI perception – the ability for machines to process and interpret sensory information – alongside advancements in language modelling, robotics, and collaborative AI agents.

Meta stated its goal involves creating machines “that are able to acquire, process, and interpret sensory information about the world around us and are able to use this information to make decisions with human-like intelligence and speed.”

The five new releases represent diverse but interconnected efforts towards achieving this ambitious goal.

Perception Encoder: Meta sharpens the ‘vision’ of AI

Central to the new releases is the Perception Encoder, described as a large-scale vision encoder designed to excel across various image and video tasks.

Vision encoders function as the “eyes” for AI systems, allowing them to understand visual data.

Meta highlights the increasing challenge of building encoders that meet the demands of advanced AI, requiring capabilities that bridge vision and language, handle both images and videos effectively, and remain robust under challenging conditions, including potential adversarial attacks.

The ideal encoder, according to Meta, should recognise a wide array of concepts while distinguishing subtle details—citing examples like spotting “a stingray burrowed under the sea floor, identifying a tiny goldfinch in the background of an image, or catching a scampering agouti on a night vision wildlife camera.”

Meta claims the Perception Encoder achieves “exceptional performance on image and video zero-shot classification and retrieval, surpassing all existing open source and proprietary models for such tasks.”

Furthermore, its perceptual strengths reportedly translate well to language tasks. 

When aligned with a large language model (LLM), the encoder is said to outperform other vision encoders in areas like visual question answering (VQA), captioning, document understanding, and grounding (linking text to specific image regions). It also reportedly boosts performance on tasks traditionally difficult for LLMs, such as understanding spatial relationships (e.g., “if one object is behind another”) or camera movement relative to an object.

“As Perception Encoder begins to be integrated into new applications, we’re excited to see how its advanced vision capabilities will enable even more capable AI systems,” Meta said.

Perception Language Model (PLM): Open research in vision-language

Complementing the encoder is the Perception Language Model (PLM), an open and reproducible vision-language model aimed at complex visual recognition tasks. 

PLM was trained using large-scale synthetic data combined with open vision-language datasets, explicitly without distilling knowledge from external proprietary models.

Recognising gaps in existing video understanding data, the FAIR team collected 2.5 million new, human-labelled samples focused on fine-grained video question answering and spatio-temporal captioning. Meta claims this forms the “largest dataset of its kind to date.”

PLM is offered in 1, 3, and 8 billion parameter versions, catering to academic research needs requiring transparency.

Alongside the models, Meta is releasing PLM-VideoBench, a new benchmark specifically designed to test capabilities often missed by existing benchmarks, namely “fine-grained activity understanding and spatiotemporally grounded reasoning.”

Meta hopes the combination of open models, the large dataset, and the challenging benchmark will empower the open-source community.

Meta Locate 3D: Giving robots situational awareness

Bridging the gap between language commands and physical action is Meta Locate 3D. This end-to-end model aims to allow robots to accurately localise objects in a 3D environment based on open-vocabulary natural language queries.

Meta Locate 3D processes 3D point clouds directly from RGB-D sensors (like those found on some robots or depth-sensing cameras). Given a textual prompt, such as “flower vase near TV console,” the system considers spatial relationships and context to pinpoint the correct object instance, distinguishing it from, say, a “vase on the table.”

The system comprises three main parts: a preprocessing step converting 2D features to 3D featurised point clouds; the 3D-JEPA encoder (a pretrained model creating a contextualised 3D world representation); and the Locate 3D decoder, which takes the 3D representation and the language query to output bounding boxes and masks for the specified objects.

Alongside the model, Meta is releasing a substantial new dataset for object localisation based on referring expressions. It includes 130,000 language annotations across 1,346 scenes from the ARKitScenes, ScanNet, and ScanNet++ datasets, effectively doubling existing annotated data in this area.

Meta sees this technology as crucial for developing more capable robotic systems, including its own PARTNR robot project, enabling more natural human-robot interaction and collaboration.

Dynamic Byte Latent Transformer: Efficient and robust language modelling

Following research published in late 2024, Meta is now releasing the model weights for its 8-billion parameter Dynamic Byte Latent Transformer.

This architecture represents a shift away from traditional tokenisation-based language models, operating instead at the byte level. Meta claims this approach achieves comparable performance at scale while offering significant improvements in inference efficiency and robustness.

Traditional LLMs break text into ‘tokens’, which can struggle with misspellings, novel words, or adversarial inputs. Byte-level models process raw bytes, potentially offering greater resilience.

Meta reports that the Dynamic Byte Latent Transformer “outperforms tokeniser-based models across various tasks, with an average robustness advantage of +7 points (on perturbed HellaSwag), and reaching as high as +55 points on tasks from the CUTE token-understanding benchmark.”

By releasing the weights alongside the previously shared codebase, Meta encourages the research community to explore this alternative approach to language modelling.

Collaborative Reasoner: Meta advances socially-intelligent AI agents

The final release, Collaborative Reasoner, tackles the complex challenge of creating AI agents that can effectively collaborate with humans or other AIs.

Meta notes that human collaboration often yields superior results, and aims to imbue AI with similar capabilities for tasks like helping with homework or job interview preparation.

Such collaboration requires not just problem-solving but also social skills like communication, empathy, providing feedback, and understanding others’ mental states (theory-of-mind), often unfolding over multiple conversational turns.

Current LLM training and evaluation methods often neglect these social and collaborative aspects. Furthermore, collecting relevant conversational data is expensive and difficult.

Collaborative Reasoner provides a framework to evaluate and enhance these skills. It includes goal-oriented tasks requiring multi-step reasoning achieved through conversation between two agents. The framework tests abilities like disagreeing constructively, persuading a partner, and reaching a shared best solution.

Meta’s evaluations revealed that current models struggle to consistently leverage collaboration for better outcomes. To address this, they propose a self-improvement technique using synthetic interaction data where an LLM agent collaborates with itself.

Generating this data at scale is enabled by a new high-performance model serving engine called Matrix. Using this approach on maths, scientific, and social reasoning tasks reportedly yielded improvements of up to 29.4% compared to the standard ‘chain-of-thought’ performance of a single LLM.

By open-sourcing the data generation and modelling pipeline, Meta aims to foster further research into creating truly “social agents that can partner with humans and other agents.”

These five releases collectively underscore Meta’s continued heavy investment in fundamental AI research, particularly focusing on building blocks for machines that can perceive, understand, and interact with the world in more human-like ways. 

See also: Meta will train AI models using EU user data

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Meta FAIR advances human-like AI with five major releases appeared first on AI News.

The announcement follows the recent launch of Meta AI features in Europe and aims to enhance the capabilities and cultural relevance of its AI systems for the region’s diverse population.   

In a statement, Meta wrote: “Today, we’re announcing our plans to train AI at Meta using public content – like public posts and comments – shared by adults on our products in the EU.

“People’s interactions with Meta AI – like questions and queries – will also be used to train and improve our models.”

Starting this week, users of Meta’s platforms (including Facebook, Instagram, WhatsApp, and Messenger) within the EU will receive notifications explaining the data usage. These notifications, delivered both in-app and via email, will detail the types of public data involved and link to an objection form.

“We have made this objection form easy to find, read, and use, and we’ll honor all objection forms we have already received, as well as newly submitted ones,” Meta explained.

Meta explicitly clarified that certain data types remain off-limits for AI training purposes.

The company says it will not “use people’s private messages with friends and family” to train its generative AI models. Furthermore, public data associated with accounts belonging to users under the age of 18 in the EU will not be included in the training datasets.

Meta wants to build AI tools designed for EU users

Meta positions this initiative as a necessary step towards creating AI tools designed for EU users. Meta launched its AI chatbot functionality across its messaging apps in Europe last month, framing this data usage as the next phase in improving the service.

“We believe we have a responsibility to build AI that’s not just available to Europeans, but is actually built for them,” the company explained. 

“That means everything from dialects and colloquialisms, to hyper-local knowledge and the distinct ways different countries use humor and sarcasm on our products.”

This becomes increasingly pertinent as AI models evolve with multi-modal capabilities spanning text, voice, video, and imagery.   

Meta also situated its actions in the EU within the broader industry landscape, pointing out that training AI on user data is common practice.

“It’s important to note that the kind of AI training we’re doing is not unique to Meta, nor will it be unique to Europe,” the statement reads. 

“We’re following the example set by others including Google and OpenAI, both of which have already used data from European users to train their AI models.”

Meta further claimed its approach surpasses others in openness, stating, “We’re proud that our approach is more transparent than many of our industry counterparts.”   

Regarding regulatory compliance, Meta referenced prior engagement with regulators, including a delay initiated last year while awaiting clarification on legal requirements. The company also cited a favourable opinion from the European Data Protection Board (EDPB) in December 2024.

“We welcome the opinion provided by the EDPB in December, which affirmed that our original approach met our legal obligations,” wrote Meta.

Broader concerns over AI training data

While Meta presents its approach in the EU as transparent and compliant, the practice of using vast swathes of public user data from social media platforms to train large language models (LLMs) and generative AI continues to raise significant concerns among privacy advocates.

Firstly, the definition of “public” data can be contentious. Content shared publicly on platforms like Facebook or Instagram may not have been posted with the expectation that it would become raw material for training commercial AI systems capable of generating entirely new content or insights. Users might share personal anecdotes, opinions, or creative works publicly within their perceived community, without envisaging its large-scale, automated analysis and repurposing by the platform owner.

Secondly, the effectiveness and fairness of an “opt-out” system versus an “opt-in” system remain debatable. Placing the onus on users to actively object, often after receiving notifications buried amongst countless others, raises questions about informed consent. Many users may not see, understand, or act upon the notification, potentially leading to their data being used by default rather than explicit permission.

Thirdly, the issue of inherent bias looms large. Social media platforms reflect and sometimes amplify societal biases, including racism, sexism, and misinformation. AI models trained on this data risk learning, replicating, and even scaling these biases. While companies employ filtering and fine-tuning techniques, eradicating bias absorbed from billions of data points is an immense challenge. An AI trained on European public data needs careful curation to avoid perpetuating stereotypes or harmful generalisations about the very cultures it aims to understand.   

Furthermore, questions surrounding copyright and intellectual property persist. Public posts often contain original text, images, and videos created by users. Using this content to train commercial AI models, which may then generate competing content or derive value from it, enters murky legal territory regarding ownership and fair compensation—issues currently being contested in courts worldwide involving various AI developers.

Finally, while Meta highlights its transparency relative to competitors, the actual mechanisms of data selection, filtering, and its specific impact on model behaviour often remain opaque. Truly meaningful transparency would involve deeper insights into how specific data influences AI outputs and the safeguards in place to prevent misuse or unintended consequences.

The approach taken by Meta in the EU underscores the immense value technology giants place on user-generated content as fuel for the burgeoning AI economy. As these practices become more widespread, the debate surrounding data privacy, informed consent, algorithmic bias, and the ethical responsibilities of AI developers will undoubtedly intensify across Europe and beyond.

(Photo by Julio Lopez)

See also: Apple AI stresses privacy with synthetic and anonymised data

Want to learn more about AI and big data from industry leaders? Check out AI & Big Data Expo taking place in Amsterdam, California, and London. The comprehensive event is co-located with other leading events including Intelligent Automation Conference, BlockX, Digital Transformation Week, and Cyber Security & Cloud Expo.

Explore other upcoming enterprise technology events and webinars powered by TechForge here.

The post Meta will train AI models using EU user data appeared first on AI News.