7 OCTOBER | LONDON 2024
SEPTEMBER 12TH - 14TH
The O2, LONDON
AI & DeepTech
Your weekly newsletter on cutting-edge innovations in AI, biotech & quantum
The week's developments in AI, quantum, and biotech, explained | 01.05.24
If you’re enjoying this briefing, sign up here — and share it with a friend. Reading time: ~4 minutes
This week, French startup Klineo raised €2M to connect doctors and patients with cancer clinical trials through their AI matchmaking platform. Plus researchers have improved radiotherapy dose prediction through Deep Evidential Learning, with huge promise for clinical applications. Will the future of healthcare be stitched together by AI?
Meanwhile, in robotics, this week saw both progress and setbacks. On one hand, robots are learning to fall safely (see Boston Dynamics new Atlas — it’s incredible). On the other, they’re still significantly slower than animals, and researchers may have found why.
Keep reading for how researchers are building artificial life (through artificial cells), universal compute, and the future of AI coffee — alongside the coolest AI tools to optimise, enhance and organise…
— Charlie and the Research & Intelligence Team
P.S You are now able to sign up for the super early bird offer (75% off) to the CogX AI Summit in London on October 7th here.
Innovation and Releases
💉 Startup raises €2M to make cancer clinical trials more accessible via AI: Klineo helps matchmake patients and doctors with relevant trials, focusing on skin, breast, and lymphoma cancers. The company plans to extend its coverage to all cancer types by year's end.
🚀Boeing's Starliner spacecraft will launch first crewed mission on May 6. Astronauts will travel to the International Space Station, where they will spend a week. During their stay, the Starliner will remain docked at the ISS, and it will be used for their return journey to Earth.
🤖Humanoid robots are learning to fall well, Boston Dynamics and Agility recognising falls are inevitable in the real-world. The approach will aid manoeuvring in uneven environments and also helps in understanding and improving robotic performance through trials and errors.
💻Universal compute added to France’s AI arsenal: Startup FlexAI has emerged with €28.5 million to tackle high compute demands. Their ‘universal compute service’ will simplify the development and training of AI on diverse hardware without intricate code changes.
☕️The future of coffee is ‘AI-conic’ according to Helsinki’s Kaffa Roastery with the release of their "AI-conic" coffee blend, unique AI-generated combination of four different beans: 40% Brazil, 25% each from Colombia and Guatemala, and 10% from Ethiopia.
AI Tools of the Week
🚀Optimise: TheFastest.ai is a performance benchmarking tool for LLMs tracking metrics like Time To First Token (TTFT) and Tokens Per Second (TPS).
🛠️Enhance: Albus is an AI Slack assistant that creates generates strategy ideas, analyses market trends, crafts reports and produces marketing content.
✍️Organise: Saner.AI is an AI-powered personal knowledge management tool crafted to streamline information organisation and utilisation.
Latest Research
🏥Researchers enhanced radiotherapy dose prediction using Deep Learning, showing its ability to distinguish between knowledge-based uncertainties and those from noisy data. This improves prediction accuracy, making it highly valuable for medical applications.
🤖Why are robots still slower than animals? A recent study found that while individual robot parts could outperform biological ones, their overall integration and performance lag behind, suggesting engineers should learn from biology to improve system coordination.
🚗This automatic AI car can drive with confidence, as developers leveraged a Bayesian Neural Network to manage uncertainties. This system adapts well across various simulated tracks and uses prediction confidence to determine when manual intervention is needed.
🧬The beginning of artificial life? Researchers have successfully created artificial cells that mirror the behaviour of living cells. The team manipulated DNA and proteins to engineer cells that respond dynamically to environmental changes.
In case you missed it
An interview with (former CogX speaker) NVIDIA’s Jensen Huang and his 2 trillion dollar company powering tomorrow’s AI:
✍️ Enjoying this newsletter? Subscribe to our free weekly briefings on Preparing for AI, Cinema & AI, The Race to Net Zero, AI & DeepTech and The Future of Work.
🚀 You are now able to sign up for the super early bird offer (75% off) to the CogX AI Summit in London on October 7th here.