- Markets - Around The World
- Posts
- ChatGPT is 'Rewiring' Our Brains
ChatGPT is 'Rewiring' Our Brains
& CloudFlare's Web Paywall
Is ChatGPT Making Us Dumber—or Just Changing How We Think?
ChatGPT might not be dumbing us down— but it’s definitely rewiring how we think. Just like calculators shifted our need to do long division manually, AI shifts our mental energy from rote tasks to creative, critical thinking.
MIT conducted a controlled experiment, where participants were split into two groups—one answered questions solo, the other used ChatGPT. The AI group performed better on both factual and creative tasks and reported feeling more confident in their answers.
But here's the nuance: people who treated ChatGPT like a collaborator—not just a copy-paste machine—got the most out of it. It boosted performance and engagement when used thoughtfully. The takeaway? AI doesn’t rot your brain. Like any tool, it amplifies your habits. Use it passively, and your thinking gets lazier. Use it intentionally, and it becomes a powerful partner in sharper, deeper thinking.
Meet the Paywall: Cloudflare Just Put Up a Toll Booth on the Internet
What if every time an AI bot visited a website, it had to pay for the privilege? Cloudflare just flipped that switch—charging AI crawlers for every scrape and giving websites the power to block freeloaders.
Behind the scenes, Cloudflare is tracking bot traffic and bills them based on usage, forcing companies like OpenAI, Google, and Anthropic to negotiate data licenses or risk being shut out. It's a response to growing frustration from publishers whose content has been freely scraped to train large language models.
With Cloudflare guarding nearly 20% of the web, this could rewrite the rules of how the internet fuels AI. Will this toll road slow down the AI race—or simply make it more exclusive?
A Team of AIs May Be Smarter Than One Giant One
Sakana AI, founded by ex-Google DeepMind researchers in Tokyo, has developed TreeQuest, a novel approach that uses teams of smaller AIs instead of one massive model to solve complex tasks.
Each model tackles parts of the problem independently, then votes on the best answer — like a panel of experts. This “multi-model swarm” approach outperformed larger single models by 30% on tasks like logic and math, while being far more efficient.
Rather than scaling up endlessly, TreeQuest shows that coordination and collaboration might be the smarter way forward in AI — potentially changing how we build intelligent systems entirely.