3 things direct from the future

Edition 67

Once every 2 weeks I will deliver “3 things direct from the future”. A 2 minute read that will always give you:

  • one thing that can help,
  • one thing to be wary of, and
  • one thing to amaze.

If this sounds interesting to you then please subscribe.

Subscribe

1. One thing that helps

I Think Therefore I Text

ALS patients have begun to receive an implant that will help them communicate using thought. Amazingly, sending texts and online shopping just by thinking is now possible, thanks to Synchron’s wireless brain-computer interface.

Stentrode

Four patients in Australia have already received this “stentrode” with no adverse effects, and on July 6, they made their first implant on a patient in the US (as apparently here in Australia we have less cautious rules around this stuff!). The stentrode is a cylindrical hollow mesh wire device that is inserted into a blood vessel in the motor cortex. Only a small incision in the neck is needed for the catheter that delivers the stentrode. OK, it still doesn’t sound super fun, but much better than other methods where they cut a portion of the skull – and this one only takes a few minutes.

The stentrode is then connected to a small computer on the patient’s chest. Once connected, it reads nerve signals in the brain which are amplified and sent to an external device via Bluetooth. Over time, the signal strength gets clearer and the software can match brain data patterns to a specific goal.

The tech is still in trial mode and cannot yet translate a whole sentence. It currently operates letter-by-letter with a “yes or no” command process. Regardless, this is a huge step towards really helping people with communication challenges and is an astounding technology.

 

2. One to be wary of 

Dr. Toxin

292936631 1085130219047899 4246264896648582867 n

Way back in Edition 36, we talked about how AI is helping researchers come up with new drugs and antibiotics. What happens if scientists turn all “Dr Evil” and flip the AI from good to bad? The result is 40,000 biowarfare agents.

A group of researchers that released a paper titled “Dual use of artificial-intelligence-inspired drug discovery” (they really have a way with names) have recognised that within the technology they are building, is the potential for catastrophic misuse.

“The thought had never previously struck us. Our work is rooted in building machine learning models for therapeutic and toxic targets to better assist in the design of new molecules for drug discovery. We have spent decades using computers and AI to improve human health—not to degrade it.”

Collaborations Pharmaceuticals designed MegaSyn, a molecule generator with the goal of finding compounds to be used in new drugs. However, when the researchers flipped their goal into looking for toxic compounds, the MegaSyn came up with 40,000 toxic substances in just six hours! These could be more deadly than toxins we already know about.

“By going as close as we dared, we have still crossed a grey moral boundary, demonstrating that it is possible to design virtual potential toxic molecules without much in the way of effort, time or computational resources,” they write. “We can easily erase the thousands of molecules we created, but we cannot delete the knowledge of how to recreate them.” 

Have we just opened a Pandora’s box?

3. One to amaze

Human Neuron Chips

 

Our brain is a powerful computer that needs no external power source. Australian start-up Cortical Labs is building computer chips using biological neurons from mice and people. The hope is to drastically reduce the power required to run AI systems.

They are combining neurons with traditional silicon chips to harness the advantages of both. There are two kinds of neurons used for this project: one is extracted from mouse embryos, while the other comes from turning human skin cells into stem cells then into neurons. This paves the way for biological neural networks that can solve problems in unfamiliar situations. Along with power efficiency, neuron chips may allow for faster computing with less energy wasted. 

Cortical Lab’s chip is still at the beginning of its development and has less processing power than a dragonfly brain (an interesting comparison). Their goal for now is to train it to be able to play a game of “Pong” – similar to how DeepMind tested its algorithms way back in 2013. I wonder whether silicon or neuron chips will win when pitted against each other.  Sort of like VHS vs Beta maybe?

Have a great week.

Daniel J McKinnon

Connect on LinkedIn

Subscribe

Subscribe

Don’t Stop Here

More To Explore