Whole Brain Emulation: A Giant Step for Neuroscience

Machine interfaces today can link up brains to play tetris together. Like it’s not hard enough to find a place for the L-shaped block without another cerebrum trying to overrule you. 

Let’s go farther: What if we could create a digital replica of your brain and upload and download it like a piece of software?

This feat, aka whole brain emulation (WBE), is still decades, perhaps more than a century away. Outside of the pure science challenge, it could make us confront some of the most daunting questions about what it means to be human, and where man ends and machine begins.

Whole Brain Emulation 101

The term “whole brain emulation” might sound new, but chances are you’ve seen it across popular fiction. Several episodes of Netflix’s Black Mirror have explored potential manifestations of brain emulating and mind uploading, including:

• White Christmas: Digital clones of people (“cookies”) are used to carry out a variety of jobs, and one man’s clone gets him into some legal trouble.
• San Junipero: A simulated reality allows individuals to live forever.
• Callister: Digital clones that think, feel, and act like the humans whose DNA they were pulled from become trapped in a simulated reality.
• Hang the DJ: A mind upload is tested for romantic compatibility with partners in a different simulated reality.
• Be right back: A virtual copy of deceased individuals allows their loved ones to never have to let go.

So how do we make those stories a reality? In a 2008 whitepaper, futurists Nick Bostrom and Anders Sandberg of the University of Oxford’s Future of Humanity Institute published the first roadmap for WBE.

They identified three core components: 1) scanning a brain 2) interpreting the brain data and building a software model and 3) simulating this model “so faithful[ly] to the original that, when run on appropriate hardware, it will behave in essentially the same way as the original brain.” It’s closely tied to the concepts of “mind uploading” and “downloading”–but even that phrasing needs some unpacking.

• The “brain” is that biological mash of neurons and synapses that makes you think, feel, and experience. The “mind” is more ambiguous: Some view it as separate from the brain, others as intrinsically woven together. We’ll leave that debate to philosophy 101 courses.

Bostrom and Sandberg laid out a series of benchmarks they believed WBE technology would follow. These include functional brain emulation (eg- showing awake versus asleep), species-generic emulation (a range of typical human behavior and learning capacity), then individual brain emulation (distinct to one person’s brain).

• But due to “the complexities and conceptual issues of consciousness,” they stopped just short of assessing the final benchmarks: mind, social role-fit, and personal identity emulation.

Let’s go deeper

Getting your head into the cloud is not just a computer and neuroscience question. Humans have been philosophizing about what the mind is and how it relates to the brain for centuries.

Mitch Green, professor of philosophy at UConn, described one hypothetical end game for WBE: You enter a booth, scan and upload your mind to a server, then download it into a robot body. Without going anywhere, you spend an afternoon wine tasting in Napa and skiing in Austria.

Right now, scenarios like this create more questions than they answer. Can the mind persist without its biological substrate (aka its brain)? If your mind is in an android, do you exist in two places at once? Will you need two alibis when the police come knocking?

Now, as great as android uploads would be, whether the individual would survive is another question (complicated even further by questions of faith and religion), according to Susan Schneider, philosophy professor and director of the AI, Mind and Society Group at UConn. Schneider noted she finds it “crazy” that consciousness could be transferred to another location.

• If you want to be even more confused, consider this as well: If you slowly replace your brain piece-by-piece with computer chips, is your mind still the same at the end, or would you lose yourself in the process, Schneider asked.
• Even on the assumption the mind could be abstracted and transferred, there isn’t a clear consensus of what success would look like.

The mind consists of cognitive, affective, and experiential qualities, and all three contribute to what we understand as the human experience, according to Green. If your mind takes that trip to Napa Valley, does the robot actually taste the 2007 Pinot Noir? If the answer is yes, whose mind is sensing and experiencing it–yours, or the robot’s? 

We don’t like leaving you with more questions than answers, so let’s get back to the science.

Insane in the brain

Neuroscience researchers understand how neurons, synapses, and electrical signals make the brain work. Higher-level cognition and real-time observation, a little less so.

In his day job, Kenneth Hayworth, president and co-founder of the Brain Preservation Foundation, works on mapping insect brains, with plans to graduate to mouse brains soon. For scale: A fruit fly or ant has about 250,000 neurons; a house mouse, 71 million. The human brain…86 billion. So yeah, mapping a human brain (a crucial step for WBE) is going to be hard.

• Randal Koene, chairman of Carboncopies (a nonprofit supporting the field of WBE) said he would be “astonished” if scientists had even a simple mind upload of a mouse within 20 years, but “equally astonished” if the human brain isn’t emulated in the next century.
• Bostrom and Sandberg also said WBE “should be possible before mid-century,” though more conservative estimates put this feat more than a century away.

Don’t let the news cycle confuse you: Just because someone is working on brains+computers today does not mean it’s whole brain emulation.

Earlier this week, Elon Musk’s secretive brain-machine interface company, Neuralink, unveiled flexible brain “threads.” All you need is a few holes drilled in your skull to implant strings finer than a human hair into your brain. The ultimate goal? To “achieve a symbiosis with artificial intelligence.” That’ll be a big nope from us, thank you very much…

What Neuralink and related companies, such as Kernel, are working on is about as cutting edge as it gets (though rats whose skulls were drilled into to get here might disagree). But the technology is focused on linking up mind/brain and computer, not creating a brain emulation that sits on hardware. 

• Keep in mind, this work is complementary to WBE. Every breakthrough in neuroscience or brain-computer interfaces helps us understand the human brain, how it will relate to machines, and what we still need to decrypt.

The nascent ecosystem still has a lot of growth ahead. It’s not a field for quick money or products of easily defined value, so driving support outside of fundamental research can be a challenge, Koene said.

That’s a lot of heavy stuff, so here’s the breakdown

The promise: The ability to scan your brain and create a digital model just like the original. 

The roadblocks: Scientists are still years away from mapping the 71 million neurons in a mouse brain. The human brain has…86 billion. And we aren’t even sure if the mind can be emulated. 

The projected timeline: Whole brain emulation is at least several decades, if not more than a century, away. 

The major players: Neuroscience research labs across the world are demystifying the human brain, one small step at a time. Companies like Musk’s Neuralink are working on brain-computer interfaces, a complimentary tech that will inform future emulation.