This talk began as a set of essays on Medium about existential risk
Existential risk covers a broad range of topics, but Rob focused most on synthetic biology risk
What is synthetic biology (SynBio)?
The artificial manipulation of DNA to make novel/non -natural “critters” that have unique powers and capabilities
It has potential for good – SynBio will probably end the organ shortage and reduce a huge about of suffering
BUT – there can be a dark side to its application if it falls into malevolent (or incompetent) hands
What could go wrong?
Let’s use an example:
To modify something like the bird flu virus to make it a highly lethal pathogen a decade ago, you would have had to be an incredibly credentialed and well-supplied scientist (but we’re moving further and further away from this reality)
“What’s interesting about synthetic biology and also potentially dangerous is it’s moving forward at a rate that’s far faster than Moore’s Law… the rate at which we’re able to synthesize base pairs and the rate at which breakthroughs like CRISPR are coming onto the gene editing scene goes far beyond the rate of development we saw with computing”
There are things world-class synthetic biology labs couldn’t possibly do today that will probably be very easy to do in a community college or high school biology lab 20 years from now
We sequenced a single haploid genome over a 13 year span – it cost $3 billion
NOW – a single haploid genome can be sequenced <$1k by someone like a well-trained high school grad
What is gene sequencing?
It’s just a fancy word for “reading”
This is what 23andMe does – they read the base pairs of your DNA to understand how your genes are laid out and what mutations you have
Let’s Get Dark
Take the H5N1 flu strain
60% of the people who contract it die
But it’s not very contagious – it’s killed <50 people since 2015
In 2011, two separate research teams modified the H5N1 genome to make it just as deadly AND wildly contagious
Here’s the kicker – Rob estimates that the instructions to modify the H5N1 genome to make it highly contagious could fit on ONE single Post-it note
Since 2011, CRISPR has come along, making it radically easier to edit genomes (it’s also reduced the cost of reading genes)
Just imagine the year is 2026:
A viralogist, for a PhD thesis, makes something as deadly as Ebola, as contagious as chickenpox, AND it incubates in the human body for a full year before showing any symptoms (and the modifications are contained within a single data file)
The whole world could be inflicted before the first sign of trouble
Think – what if the university gets hacked? (it happens all the time)
Then the data file spreads to the internet’s darkest corners – who knows where that might lead?
As time goes on, it’ll just get easier and easier to actually create the “critter” on something like a desktop DNA synthesizer
“Once a genius has made a data file, any idiot can distribute it worldwide and eventually, when the printers exist, print it”
Technology is the Force Multiplier
There were 323 mass shootings in the U.S. last year (the “bad actors” are out there)
But realize – the force multiplier is technology. These bad actors tend to go for the most lethal items they can get their hands on.
In China, for example – the most lethal items people can get their hands on are hammers and knives
But in the U.S. – we have guns
Imagine what’s going to happen when deadly synthetic biology weapons become widely available?
“The worse the technology gets, when you have one of these people who detonate, the more casualties there will be”
At this point, there are probably a handful of borderline geniuses that could make a “doomsday bug”
But as technology improves – THOUSANDS of life science researchers/students/grad students will have this capability
And realize – “When you start getting into groups that big, some tiny percentage of a very large group will reliably do something awful eventually”
Look at the Vegas killer – he killed 58 people
He only stopped because he ran out of bullets – these types of deranged people are out there
How can we prevent these doomsday scenarios? – The “Immune System”
Imagine ubiquitous pathogen detectors/sensors
Today they’re primitive, but with a little bit of R&D muscle, we could push them onto their own Moore’s Law Curve where they become as widespread as smoke detectors and as networked as smartphones
How do they work?
Simply – They continuously sample air molecules
These already exist at airports, but they need to be much more ubiquitous, distributed, and networked (they need to be in constant dialogue with a central repository of pathogens)
There’s no reason they couldn’t one day be part of a smartphone
A distributed biomanufacturing base (printer technology for vaccines and medicines)
Imagine being able to print vaccines in your own home at a moment’s notice
“We need to start worrying about all this now… we can’t start worrying about this in 30 years. Think of all the billions of dollars spent on digital security and the number of viruses that still get through. The human body is so much more complex than Windows OS.”
We NEED to start imagining these worst-case scenarios NOW and how we can go about preventing them
“These extensional outcomes are OPTIONAL not destiny”