ExoBrain weekly AI news

14th March 2025: Manus agent hype, copyright battles pit tech against creators, and Gemini's native image mode arrives

Joel Miller
March 14, 2025

Welcome to our weekly email newsletter, a combination of thematic insights from the founders at ExoBrain, and a broader news roundup from our AI platform Exo…

Themes this week:

Manus, the hyped Chinese AI agent built on Claude and Qwen with notable product innovations
The copyright battle between tech giants and creators shaping AI's regulatory future
Google's Gemini 2.0 Flash and its seamless native image editing capabilities

Manus agent hype

This week, Manus, a new AI agent from Chinese startup Monica.im, "went viral" capturing the attention of numerous AI commentators and polarising camps into the breathless and the sceptical. Marketed cleverly through scarce invite codes (ExoBrain is still waiting, somewhere in a supposed 2 million strong queue) and influencer-driven promotion, Manus rapidly dominated the headlines, labelled by some as another “DeepSeek moment” for China, or even a glimpse of AGI.

Yet, Manus isn't ground-breaking in technological innovation terms in the same way R1 was. It's essentially an interface built around Anthropic and Alibaba’s existing Claude and Qwen model, augmented with web browsing and command-line tools. Its strength lies not in pioneering new AI techniques, but in creatively enhancing the conversational interface users already know well.

Interestingly, Manus predominantly operates by executing code to complete varied tasks, an approach inspired by a 2024 research project known as CodeAct, where agents use executable Python code to unify their action space. CodeAct integrates a Python interpreter to dynamically adjust actions based on observations. While you’re waiting for your invite code, the Manus provide a series of recordings where you can see this code heavy process in action. The product is not without significant flaws. Users quickly exposed Manus’s basic security architecture, prompting the agent to share parts of its source code, revealing internal details and its reliance on Claude 3.7 Sonnet. Moreover, despite impressive demonstrations of tasks such as property research and financial analysis, many users found real-world performance inconsistent or exaggerated.

The Manus affair reveals important lessons about the types of innovation users actually want. While significant efforts go into refining chatbot windows or conversely experimenting with radical new interfaces such as Flora's "intelligent canvas," there's a noticeable gap. Users often resist radically new paradigms due to the comfort of chat. Manus addresses this issue, materially extending the familiar chat model by embedding autonomous, agent-like features. The hype and criticism surrounding Manus highlight genuine user interest for conversational AI that moves beyond simple exchanges to achieve meaningful automation and practical productivity boosts. Manus’s approach, extending rather than reinventing conversational interfaces, offers a potential template for future UX innovation.

As AI policy commentator Dean Ball stated; “The Western AGI obsession makes us want to conceptualise [it] as one godlike model that can do everything, and we implicitly dismiss product engineering and practical applications. You see that reflected in public policy, which is obsessed with big models, giant datacentres, and similar infrastructure. Those are the only things we seem to take seriously and value.” Could it be that clever product design and not scale might unshackle AI?

Takeaways: Looking forward, Manus demonstrates the value of AI product engineering that explores the space between familiarity and novelty. As we reach the limits of refining traditional chatbots and encounter resistance adopting future paradigms, the "conversational-plus" space Manus occupies might become the centre of gravity for the next AI developments. Products like Google’s NotebookLM, Cursor for development, and now Manus suggest the next wave in AI won't just rely on more powerful models, but rather smarter ways of using models to deliver agentic capabilities through interfaces we intuitively trust. Manus isn’t revolutionary, but its success signals clear demand for conversational AI that combines familiarity with genuine autonomy. Expect this product space—intuitive yet powerful conversational agents—to become a new frontier of practical innovation.

Copyright battles pit tech against creators

OpenAI and Google have both submitted policy proposals to US government arguing that AI systems should be allowed to train on copyrighted materials without permission. In a time of growing geo-political confrontation, they frame this as essential for national security and economic competitiveness against China.

The pushback has been swift and strong. In the UK, where similar policies are under consideration, the Intellectual Property Office received over 13,000 responses to its consultation. High-profile artists including Elton John and Paul McCartney have protested by releasing a "silent album" on Spotify titled "The British government must not legalise music theft to benefit AI companies."

OpenAI is specifically asking the US government to protect the "freedom to learn" - codifying AI training as fair use regardless of copyright status. Their proposal offers no clear compensation framework for creators whose works train their systems. They argue that training is transformative and distinct from reproduction, suggesting that while their models learn from copyrighted works, they don't produce direct copies. This distinction is highly contested by rights holders who see their creative expressions being absorbed and potentially recreated. While tech companies argue unrestricted access to training data is vital for innovation, creators and publishers fear their work will be exploited without fair compensation. The UK is considering an "opt-out" approach allowing training on legally accessible materials unless rightsholders expressly refuse.

This clash represents more than a technical legal dispute. It highlights fundamental questions about the balance between technological progress and creative rights in the AI age.

Takeaways: The tech giants will likely win this battle. Global governments appetite for AI leadership and economic growth will outweigh creator concerns. The result? AI companies securing broad training rights with minimal compensation schemes that favour large publishers. Small creators will bear the brunt, while courts struggle to enforce meaningful boundaries especially as training activity proliferates and moves far beyond the small number of labs we have today. This isn't just about copyright - it's about whether creative work retains its value in an AI-driven economy, and technological progress in this case will not be contained by traditional governance boundaries.

Gemini's native image mode arrives

Google's Gemini 2.0 Flash saw native image generation capabilities enabled this week, and finally we get to see what a truly multi-modal model can do. Unlike previous systems that relied on separate models working together, Gemini integrates everything in one. The result? Simple text commands produce remarkably accurate image, edits, images with text and iterative variations in seconds.

As shown here, a famous artwork transformed with a single instruction. No complex prompting or technical knowledge required. This represents the first time a major tech company has shipped such seamless multimodal capabilities directly to consumers.

Takeaways: This tech will make image creation and now editing accessible to everyone. Expect new creative possibilities. We're also likely to see applications we haven't even imagined yet, perhaps in education, healthcare visualisation, or real-time collaborative storytelling. The race for multimodal AI leadership has entered a new phase, with Google currently in the lead.

Weekly news roundup

This week's news highlights major infrastructure investments in AI, growing concerns about AI governance and regulation, and significant developments in AI research and hardware capabilities, particularly in Asia.

AI business news

Apple's (AAPL) Siri chief calls AI delays ugly and embarrassing, promises fixes (Shows how even tech giants are struggling to keep pace with rapid AI developments)
After DeepSeek, Chinese fund managers beat High-Flyer's path to AI (Indicates growing investment interest in Chinese AI companies)
Adobe stock drops as AI monetisation concerns weigh on investors (Highlights challenges in converting AI capabilities into revenue)
For ServiceNow, workflow means AI agents whenever possible (Shows how AI agents are becoming central to enterprise workflow automation)
UiPath looks for a path to growth with Peak agentic AI acquisition (Demonstrates consolidation in the AI automation market)