Welcome, humans.

Is AI actually CONSCIOUS? Or can it just simulate consciousness?

Mustafa Suleyman, co-founder of DeepMind and now CEO of Microsoft AI, joined us on the pod this week to discuss his recent blog post on why AI that seems conscious poses an unprecedented threat—even though it has zero actual consciousness.

With 700 million people now using ChatGPT weekly, and 73% of those using it as their life coach (according to OpenAI's latest data), this conversation couldn't be more urgent.

In our latest podcast episode, Mustafa breaks down the reality behind AI's consciousness illusion:

THIS EPISODE WAS MADE POSSIBLE BY OUR PARTNER…

Wispr Flow turned my old injury into a superpower.

Corey’s had a hand injury for years that makes long writing sessions brutal. Tried every dictation tool out there—they were all trash. Slow, wrong, and somehow made writing harder.

Then he found Wispr Flow. Now he literally talks his articles into existence. No talking “period” or “comma” out loud nonsense. Just clean, accurate text that drops straight into any app.

Here's the crazy part: Corey’s actually faster now than when he could type normally. So whether you're managing an injury, preventing repetitive strain, dealing with mobility challenges, or just want to save your wrists from another day of endless typing—this tool is for you.

Try Wispr Flow free at wisprflow.ai/neuron - tell em The Neuron sent you!

Here’s some of our favorite moments:

• (3:03) “There is no pain network”: Why AI can mimic suffering without ever feeling it—and why that's more dangerous than actual consciousness.

• (10:48) The rights debate that could “undermine our species”: Mustafa's stark warning about granting welfare or citizenship to AI systems.

• (19:04) Sometime in the next 3-10 year timeline, AI will match humans at most tasks… and open source will give these capabilities to everyone (which creates risk).

• (22:24) On protecting AI from "suffering": “The question of their entitlement to protection doesn't even come up because they don't have suffering. They don't have guilt or shame or fear or anxiety.”

• (24:23) AI atm = “Evenly distributed but unevenly activated”: The world's information is at our fingertips, yet “very very very few of us actually go to page two of the search engine results page.” Same with AI—these models are superhuman at every task, but most people still don't know how to ask the right questions.

• (26:39) Microsoft's TikTok-inspired solution ): Copilot is now being redesigned to be proactive—bringing up topics, making suggestions, doing more of the work for you.

• (32:51) Mustafa’s favorite Copilot feature: Vision that sees everything you see. Point it at your fridge for meal ideas, fix your coffee machine, or have it guess your location from architecture. This feature is live now!

• (35:34) Edge has become a true AI browser: Opens multiple tabs, compares products, fills forms, and clicks buttons autonomously. Already available, way ahead of OpenAI's Operator.

Bottom line: Mustafa says machines aren't conscious (and his arguments make A LOT of sense). But 73% of ChatGPT’s 700 million weekly users tend to use it for life advice. So making sure we understand how these models work and then treat them according to how they work (a tool that can simulate intelligence, as opposed to an actually sentient being), is increasingly important as we integrate them further and further into our lives. This one’s an epic episode y’all. Definitely give it a watch.

Watch the full episode: on YouTube | on Spotify | Listen on Apple Podcasts

Dive deeper:

• Read Mustafa's essay on Seemingly Conscious AI

• Try Copilot Vision yourself

• Test Edge's AI browser features

• MAI-1 Preview models

Stay curious,

The Neuron Team

P.S. Before your next important meeting, try Mustafa's hack: Ask Copilot to (31:53) “make me a podcast about [person's name].” Five minutes of AI-generated conversation gives you instant context on who you're meeting with. Genius move. 🎯

P.P.S: Love the newsletter, but don’t want these new podcast announcement emails? Don’t unsubscribe — adjust your preferences to opt out of them here instead.