OV038-Deep-Tech-email-header2

issue #2

Hello World,

Welcome to the second edition of Deep Thoughts.

Since we launched, the Deep Tech pod at Octopus Ventures have been busy. We’ve thrown our first Deep Thoughts community event (‘AI on the edge’), talked CRISPR-Cas gene editing with Professor Conrad Lichtenstein and Frank Massam of Nemesis Bioscience, and dived deep into deep tech, pulling together the sharpest stories and views from the cutting edge of tech.

Max-Langran-Photography-948224

If you’re a first-timer here, we are the folks who backed Evi (sold to Amazon); Swiftkey (sold to Microsoft) and Magic Pony (sold to Twitter), and here's our active portfolio.

👋🏻 connect with you all more frequently - this is definitely supposed to be a two way channel

🧠 surface some of the fascinating things our portfolio companies are working on 

📡 highlight interesting companies we have come across, as deep tech across the UK and Europe continues to accelerate at pace

💡 share with you what we are reading, chatting about across the desks and Slacking to each other so you get more of an insight into what’s getting us excited

In the last six weeks we've been...

  • Getting serious with trends in drug-resistant pathogens and the market failures in discovering new antibiotics and antimicrobials to combat this.
  • Trying to determine if Prime Editing, the newest and most precise gene editing tool available, will help to solve a range of genetic disorders including sickle cell (Nature paper here);
  • Watching the trend in Integrated Circuits (ICs) toward chiplets gather pace to trying to keep the semiconductor industry on track with Moore’s law, including Arm’s recent announcement with TSMC.
  • Excited by Facebooks acquisition of CTRL-labs! Another win for a fantastic company in the Human Machine Interface (HMI) space that we’re keeping a close eye on.
  • Not as excited as others by Googles rapidly retracted paper on (maybe) reaching quantum supremacy. We’re more into companies going straight for fault tolerant quantum computing which promises world changing advances. The consensus seems to be that the Noisy Intermediate Scale Quantum (NISQ) era we are now in wont yield many advances.
  • Reading an excellent primer on how AI models suffer from bias too - largely because how the data was trained - and how practitioners can try to avoid them. Here is the full paper.
  • Impressed to see researchers from Google Brain and UC Berkeley release an open source solution to encourage low-cost robotic reinforcement-learning experimentation. Deploying models in the real-world is essential, but typically expensive hardware meant lots of iterating was done in simulation. Can’t wait to see startups building on top of this.
  • Watching a robotic arm complete a variety of dextrous tasks by translating observed human hand/finger motion to robotic arm/finger motion. Full paper here.
  • Intrigued by this assessment, when doctors with no coding or deep learning experience developed algorithms to classify medical images. Whilst far from perfect, it shows that the development of deep learning models may not be confined to those with specific mathematical and technical expertise.
  • Enjoying this unveiling of the largest 3D printed boat in true American fanfare. Guinness world record was awarded to University of Maine Advanced Structure and Composite Centre for the same.
  • Fascinated by how artificial skin is being used to bridge the gap between humans and machines. 
  • Keeping on brand and watching an Octopus change colour whilst dreaming.

We are going to be at Slush on Nov 21-22. If you are too, say 👋🏻 hello.

We love to share, read geeky papers and always enjoy a pun-swap. So if you have something to say, email us here and if you’re just visiting, remember to sign up to receive our next edition and share this link far and wide.

Uzma

P.S. I am a biophysicist with an unrelenting thirst for learning, so almost nothing excites me more than getting a breakdown of the science underpinning the deep tech platforms we are assessing. Given my background, personally I’m super passionate about the Microbiome, Agritech, Foodtech/Alternative Proteins (foodie!) and Synthetic Biology space.

gray-line - a

Hungry for more? We’ve also been…

  • Reading some very geeky papers about a new theory of consciousness (Information Closure Theory), published by an AI company in Japan (ARAYA) and a new hypothesis for the origin of static electricity. (Yes, we still don’t know how it works!)
  • Observing more caution from researchers into voice-enabled interfaces here, proposing that a privacy-preserving layer ensures that the signal from the voice is sanitised.
  • Feeling less bullish than this author on the state of Europe’s academic institutions commercialising technology developed in their labs. We still have a long way to go to match the US leaders in this field - but it is great to see that this is high up on the agenda of these institutions.
  • Wondering if anyone in our community is developing for mixed reality? If you are, send us an email.
  • Learning how Google has open sourced BERT allowing companies to build their own state of the art question answering system in less than 30 minutes on a single cloudTPU, or a few hours on a single GPU.
  • Gripped by how OpenAI has managed to get a robot to solve a rubiks cube single-handedly. Changing the paradigm on robots’ ability to do multiple tasks concurrently. Full paper here.
gray-line - a