Good morning, AI enthusiast.

Relying on AI chatbots for tasks like writing might make things easier, but a new study suggests it comes at a cognitive cost. Research from MIT found that using tools like ChatGPT significantly reduces brain activity.

The findings indicate a direct trade-off between the convenience of AI assistance and the mental effort that helps us learn. As these tools become more integrated into our daily workflows, it raises an important question: how do we balance efficiency with the need to maintain our critical thinking skills?

In today’s AI recap:

• The cognitive cost of using AI chatbots

• Google trains Gemini on YouTube videos

• Meta’s major AI leadership acquisition

• Neuralink enables mind-controlled gaming

Your Brain on AI

The Recap: A new study from MIT using EEG headsets found that relying on AI chatbots for writing tasks significantly reduces brain activity and cognitive engagement compared to unassisted work.

Unpacked:

• Researchers tracked three groups of students writing essays: one using only their brain, one using a search engine, and one using OpenAI’s GPT-4o.

• The AI-assisted group showed the weakest neural connectivity, with brain engagement scaling down by as much as a 55% reduction compared to the unassisted group.

• The effects went beyond brain scans, as the chatbot group also had poorer recall of what they wrote and reported a lower sense of ownership over their essays.

Bottom line: These preliminary findings suggest a direct trade-off between the efficiency of AI assistance and the cognitive effort that strengthens learning. This research sparks a vital conversation about how to best integrate AI tools in education and work without short-circuiting critical thinking.

YouTube's Big Secret

The Recap: Google confirmed it's using its massive public library of YouTube videos to train its AI models, including Gemini and the new Veo 3 video generator. The move, detailed under its policy, has ignited a major debate about creator consent, compensation, and copyright.

Unpacked:

• While YouTube allows creators to opt out of having their content used for training by select third parties, there is currently no way to prevent Google from using the data for its own models.

• The practice fuels tools like the advanced Veo 3 video generator, raising concerns that creators are unknowingly helping to build systems that could one day compete with them.

• In response to intellectual property concerns, Google offers an indemnification clause for its generative AI tools, promising to assume legal responsibility if a user is challenged on copyright grounds for generated output.

Bottom line: This approach provides Google with an unparalleled data advantage in the race to build powerful generative models. It also sets a critical precedent for how user-generated content will be valued and protected as AI continues to evolve.

Where AI Experts Share Their Best Work

Join our Free AI Automation Community

Join our FREE community AI Automation Mastery — where entrepreneurs, AI builders, and AI agency owners share templates, solve problems together, and learn from each other's wins (and mistakes).

What makes our community different:

• Real peer support from people building actual AI businesses

• Complete access to download our automation library of battle-tested n8n templates

• Collaborate and problem-solve with AI experts when you get stuck

Dive into our course materials, collaborate with experienced builders, and turn automation challenges into shared wins. Join here (completely free).

Meta's Billion-Dollar AI Bet

The Recap: Meta is in advanced talks to hire influential AI investors Nat Friedman and Daniel Gross, a move that could also include a partial buyout of their multi-billion dollar venture fund.

Unpacked:

• The duo brings serious clout: Friedman is the former CEO of GitHub, and Gross is a co-founder of Safe Superintelligence (SSI) alongside former OpenAI chief scientist Ilya Sutskever.

• The deal includes a potential partial buyout of their venture fund, NFDG, which holds stakes in top AI startups like Perplexity, Character.ai, and Stripe.

• This follows Meta's recent $14.8 billion investment in Scale AI and the hiring of its founder, Alexandr Wang, signaling an aggressive push to acquire top-tier AI leadership.

Bottom line: This move shows Meta isn't just building AI, it's buying into the very ecosystem that powers it. The ongoing AI talent war now clearly extends to acquiring not just engineers, but the investors who spot the next big thing.

Neuralink's Mind Games

The Recap: Neuralink is showing remarkable progress with its brain-computer interface, as its sixth human recipient can now play video games using only their thoughts. The company also shared its ambition to help blind people see within the next year.

Unpacked:

• The sixth recipient, Rob Greiner, uses the company's Telepathy chip to move a cursor and interact with video games simply by thinking.

• Beyond motor control, Neuralink set a bold timeline, suggesting its technology could help restore sight for blind individuals in as little as six to twelve months.

• These developments are being celebrated as a way to restore function and improve quality of life for people with debilitating motor skill impairments.

Bottom line: These milestones move brain-computer interfaces from theoretical science fiction into tangible, life-altering applications. While timelines for future goals remain ambitious, the current progress demonstrates a significant leap in merging human biology with technology.

The Shortlist

OpenAI warned that its future models could significantly increase the risk of bioweapon development, anticipating that upcoming systems will be powerful enough to enable even non-experts to create biological threats.

INCOMPAS lobbied on behalf of Amazon, Google, and Microsoft for a 10-year nationwide ban on U.S. states creating their own AI regulations, arguing it would prevent a patchwork of rules and maintain America's lead in AI innovation.

Researchers found a direct trade-off between AI model accuracy and environmental impact, with models that use complex reasoning emitting up to 50 times more CO2 than more concise models for the same tasks.

China unveiled a project to remake 100 classic kung fu films using AI, including iconic titles from Bruce Lee and Jackie Chan, to reshape the visual aesthetics for contemporary audiences.

Signing off,

David, Lucas, Mitchell — The Recap editorial team