|
[Please don’t share this newsletter yet. It is in test mode until January 2025.] Donald Trump will become US president in 2025. How will that affect dual-use technology? It’s not clear yet what the implications are for Ukraine, Israel, Taiwan or indeed NATO. There will be winners and losers in each. It’s likely that technologies owned by Elon Musk will benefit. We may also see increased European spending on dual-use technologies. President Macron told European leaders last week that they must seize control of history: “Do we want to read the history written by others - the wars launched by Vladimir Putin, the US election, China’s technological or trade choices,” Macron asked. “Or do we want to write our own history? I think we have the strength to write it.” It may be that in building the strength to write history, Europeans invest more in technologies to enable that. (LINK) The Royal Navy successfully tested quantum-sensing technology on one of its P2000 ships. The Office of the Chief Technology Officer (OCTO) worked with UK quantum technology company Aquark Technologies to achieve this. Quantum sensing will become increasingly important for militaries as it can enable them to calculate their position on the earth without the need for GPS. We can expect any future conflict with a capable adversary to require non-GPS-based positioning capability. This has been demonstrated in Ukraine, with Russia adept at jamming GPS signal. This also makes clear the usefulness of small innovation teams, such as OCTO, in identifying and sourcing these technologies. (LINK) US Special Operations Command (SOCOM) updated its technical areas of interest in an announcement to include additional autonomous capabilities that could aid operators with their missions. “SOF is interested in improving Battlespace Awareness by leveraging systems where multiple Autonomous or Remote-controlled Systems (ARCS) share data in real time, allowing them to work together without constant human intervention.” This will allow for easier procurement of autonomous systems by the US DOD, and will stimulate innovation as companies look to provide better autonomous systems to US SOF. (LINK) The UK’s MOD has partnered with Surrey Satellite Technology to deliver satellite imaging via a new satellite called Juno. The £40m project will support around 200 jobs, boosting the UK’s space sector. Expected to launch in 2027, Juno will capture daytime images of the earth’s surface, building on the capabilities of Tyche, UK Space Command’s first satellite which successfully launched in August 2024. This is an important building block for UK sovereign space capability. It is also a demonstration of purchasing intent - one which will whet the appetite of other UK spacetech companies. I expect to see more European countries looking to develop this kind of sovereign space capability over the next few years. (LINK) Palantir and Anthropic have teamed up with AWS to deliver Anthropic’s “Claude AI” models to US intelligence and defence agencies. The collaboration will embed Claude’s models within Palantir’s data analytics platform, enabling a range of powerful features for agencies. For me this shows how close we are to implementing AI in government operations. It also goes to show that even Anthropic doesn’t get to do this at scale without partnering with a prime. No matter how exceptional your AI is. (LINK) TSMC has reported that its US facility is producing better results than those in Taiwan. It is apparently producing usable chips at a better rate than comparable facilities in the company’s home country of Taiwan - some 4% more are of usable quality. Although it’s unlikely that they have achieved this at the same level of nanometer (apparently the Taiwanese government has banned TSMC from making 2nm chips outside of the island) this is a major step forward for the US franchise, which had lagged behind until now. It will be unwelcome news in Taipei, which will not want the US to reduce its dependence on Taiwan for high quality chips. (LINK) Finally, MIT’s Technology Review reports that AI might be eating quantum computing’s lunch: “AI is now being applied to fundamental physics, chemistry and materials science in a way that suggests quantum computing’s purported home turf might not be so safe after all". I don’t buy this. The exponential increase in computing power as qubits are added means that quantum computing may be able to solve complex probabilistic problems such as decryption or molecule structure simulation that are simply not possible for classical computers, no matter how much AI you throw at them. (LINK) Thank you for reading this newsletter. I am working on making it better and I appreciate your feedback.
|