Four For Friday | July 19, 2024
LF132 | Two 'systems' stories this week, a drug that extends lifespan by 25% and the emergence of spatial intelligence.
Welcome to Looking Forward’s Four For Friday. Four things that have piqued my interest this week. Enjoy!
Mapping systems, systematically
This fascinating project by UK innovation agency Nesta makes the point that creating ‘systems maps’ to understand the players and key issues in a system - is slow, laborious, static and can be a bias-ridden process. The issues that are deemed as central will often be selected arbitrarily by a small group of people in a room based on gut feel. But are those the right issues for everyone? And where were things last year, and where will they be next year?
The ‘Sweet Spot’ model Nesta has developed identifies four areas that can indicate progress in an issue: investments, academic research, policy & regulations and news / press & PR.
“We’re looking for the convergence of strong research activity and significant growth in private investment alongside interest from policymakers and the media. Taken together, encouraging signals of this nature would point to a maturing, highly promising innovation or solution”
They then use big data to identify momentum in the four areas above, enabling the creation of heat maps. This allows funders, researchers and policy makers to identify progress across topic areas, and potentially also predict - or accelerate - ‘tipping points’ in the widespread distribution of certain solutions.
To fix healthcare, fix systems not people
The HBR weighs in on the need for systems change in healthcare.
Too often, the reason cited for healthcare innovation not working is put down to leadership and / or culture. This article suggests that this focus on people (and with it an inherent blame) is wrong-headed; the people are getting burned out both doing their jobs and all trying to fix the broken systems they work in.
The article notes that improving the system needs to be baked into people’s daily activities, not an afterthought.
[Effective systems operators] spend time seeking feedback and understanding challenges so they can (re)design daily operations to make frontline workers’ jobs easier and simultaneously improve the output of the system. If quality and safety are important to the work product of the organization, they can’t be ancillary concerns. They must be embedded in the work itself.
Drug extends mouse lifespan by 25% - will it do the same for us?
Researchers in an international consortium have found a drug that extends the lifespans of laboratory mice by nearly 25%, making them healthier and stronger - with an added bonus of getting fewer cancers.
These mice, nicknamed "supermodel grannies” (e.g. the one on the right in the pic above) seemed visibly younger and healthier (as well as more disease resistant). The drug targets the interleukin-11 protein, which has various beneficial properties in young animals but has been found to cause inflammation and accelerate aging in older animals.
AI meets robots: the rise of spatial intelligence
As remarkable as Large Language Models are, so far they’ve been mostly limited to the world of words and images. To get to the truly disruptive applications of AI to humanity (both positive and negative), AI is going to need to smart making sense of the physical world too. Some are calling this spatial intelligence, and it’s a hot investment topic.
The promise is that robots will be able to perceive the world and engage with it more naturally - so think truly useful humanoid robot companions rather than deterministic factory robots. A couple of the more interesting ones in this interesting segment, in my view, include Figure and ArchetypeAI
As with all things AI, a big issue is getting access to high quality data: “Unlike large language models (LLMs) such as OpenAI's Chat-GPT, which train on swaths of text-based data to spit out accurate and human-sounding responses to written or spoken queries, there is far less data to train robots' spatial intelligence.”
As AI competition heats up, the race is on for industries and use cases to surface unique and novel datasets to train these new models. Another incentive, if one was needed for traditional industries to invest in digitalization and digital transformation.
That’s all for this week. As always, feedback welcome. Feel free to share insights or links of interest.
- Stephen