"Nothing important happens without crashes".

​
So says well known economics scholar Carlota Perez.
​
Her work has contributed to our understanding of the relationship between innovation, technical and institutional change, and economic development.
​
Her framework, examining major technological disruptions and related financial crashes, is of extensive value when deploying Strategic Foresight.
​
In a nutshell, there are two phases of technological revolutions - the Installation Period, and the Deployment Period - with a Turning Point nestled between them.
​
The Installation Period is when technology comes into the market and the infrastructure is built.
​
The Deployment Period is when the technology is broadly adopted by society.
​
The critical point is that this Turning Point has historically been a major crisis, generally in the form of a financial crash...and subsequent recovery.
​
Perez found that technological revolutions are coupled with financial cycles; each cycle, which may take 50 – 60 years, consists of four phases.
​
IRRUPTION: the time of early diffusion, with intense funding of innovation in new technologies.
​
FRENZY: massive adoption and growth leads to speculative bubbles being inflated.
​
SYNERGY: the "golden age" of production and expansion, with economies of scale.
​
MATURITY: market saturation, investment moves towards the latest shiny trinket.
​
Her 2003 book "Technological Revolutions and Financial Capital" identifies five bursts of technological innovation that have occurred in industrial history over the past three centuries.
​
These five boom-and-bust cycles of technological innovation are: the industrial revolution; steam and railways; steel, electricity and heavy engineering; oil, cars and mass production; and information technology and telecommunications.
​
Technological revolutions happen when something can be done in a more simple and cost-efficient way with the goal of making more money.
​
A technological revolution taking place results in a massive replacement of one set of technologies by another occuring.
​
Every technological revolution includes changes in people, in the way organisations are run, and what new skills are needed.
​
Each technological revolution breaks old habits and challenges established norms.
​
Looking forward into the future, we can use these learnings to apply strategic foresight to what may happen with current technological innovations and potential implications for your business.
​
We posit that the Artificial Intelligence (AI) hype train is currently in the midst of the FRENZY phase.
​
Work with us to gain independent, objective, and impartial perspective on what emerging technologies could mean for your organisation today, tomorrow, and in the future.

An essential stepping stone.

​
In August, a government lab in the USA released three highly anticipated encryption algorithms that were built to withstand cyberattacks from Quantum computers.
​
A previous explainer on Quantum is here: https://lnkd.in/g8BD9GHe​
​
Researchers around the world are racing to build Quantum computers that would operate in radically different ways from ordinary computers.
​
Encryption tools rely on complex math problems that conventional computers find difficult or impossible to solve.
​
A sufficiently capable Quantum computer, though, would be able to sift through a vast number of potential solutions to these problems very quickly, thereby defeating current encryption.
​
Due to this, they could break the current encryption that provides security and privacy for just about everything we do online.
​
Quantum computing is developing rapidly.
​
Quantum technology will speed processing by an order of magnitude compared to conventional computing and impact all key economic sectors over the next 10 years.
​
The scale of the impact is such that it could potentially be worth trillions of dollars within the next decade.
​
The promise of Quantum computing is thus significant opportunity - and enormous cybersecurity threats.
​
All three finalised post-Quantum encryption standards were released by the US Department of Commerce’s National Institute of Standards and Technology (NIST)​.
​
The three are FIPS 203 (the primary standard for general encryption), FIPS 204 (the primary standard for protecting digital signatures), and FIPS 205 (intended as a backup method in case FIPS 204 proves vulnerable).
​
These post-Quantum encryption standards secure a wide range of electronic information, from confidential email messages to e-commerce transactions that propel the modern economy.
​
Though built for the future, NIST is encouraging computer system administrators to begin transitioning to the new standards, which are ready-for-use now, as soon as possible.
​
Significant investment is already taking place (previously unpacked here: https://lnkd.in/gSx2iukY).
​
Some experts predict that a device with the capability to break current encryption methods could appear within a decade, jeapordising the security and privacy of individuals, organizations and entire nations.
​
It's not all about fear and risk however.
​
Quantum technology could become a force for solving many of society’s most intractable problems, revolutionising fields from weather forecasting to fundamental physics to drug design.
​
These new standards are an essential step forward towards the broader imperative of cryptographic resilience, both now and for the future.
​
Is your organisation Quantum-ready?

The future of medicine.

​
The FDA (Food and Drug Administration) in the USA has recently cleared an innovative hemostatic gel designed to control moderate to severe bleeding in seconds.
​
Rudimentary gauze bandages are often the default response to an accident, but in the case of severe blood loss they may not be sufficient to quickly stop haemorrhage.
​
The plant-based "TRAUMAGEL" from Cresilon​ ​is designed to temporarily control the flow of blood when applied externally to a wound from a traumatic injury using proprietary hydrogel technology.
​
Hydrogels are water-based substances, a 3D network of hydrophilic polymer material with a soft consistency and moisturising properties that quickly absorbs and retains fluid.
​
TRAUMAGEL is unique in that it is supplied in a pre-filled syringe (and is thus easy to apply), requires no preparation, and is designed to stop bleeding quickly across all types of bleeds.
​
It is intended for anyone who routinely encounters traumatic wounds and need a solution to quickly and effectively stop severe bleeding.
​
These might include medical professionals and those in the armed forces.
​
The firm also makes VETIGEL – a product intended for animals, which stops the bleeding upon coming in contact with the affected part.
​
The product, debuting in 2021, can be used both internally and externally, and it is biocompatible and requires no preparation.
​
Earlier this year Cresilon was named the winner of a Gold Stevie Award in the Most Innovative Company of the Year category.
​
Unrelated, a groundbreaking study in regenerative medicine resulted in researchers manufacturing a hydrogel out of bacteria that stimulates muscle tissue regeneration.
​
The study, published in Bioactive Materials, describes how the healing power of bacteria could open new venues in the field of tissue engineering and cell therapy.
​
The research team from the DTU - Technical University of Denmark capitalised on the body’s regenerative capacity using the native bioproduction facilities which are found in bacteria.
​
Pantoan Methacrylate (PAMA), a novel hydrogel or “bactogel,” showed promising results in treating muscle injuries in rats.
​
Their findings suggest that bioactive hydrogel system made from bacteria holds “great promise for therapeutic applications such as muscle repair, regeneration, and the development of in vitro models for drug screening and disease modeling”.
​
Thus in the future plant-based hemostatic gels may be the de-facto standard for stopping bleeding, whilst bioactive hydrogels from bacteria actively regenerate muscle loss as part of musculoskeletal injury therapy.
​
What do you make of these innovations and advancements?

The main use case for Large Language Models (LLMs) is writing text nobody wanted to read.

​

Effectively the engine behind much of the generative Artificial Intelligence (AI) bandwagon which has given us the likes of ChatGPT, LLMs are lauded by salespeople and marketeers as delivering enormous business value.
​
Nonsense.
​
LLMs do not summarise text - they only shorten it; these things are not the same!
​
LLMs do not understand anything at all: there is no reasoning, and they cannot delineate between fact and fiction.
​
As such, their unchecked use can actually cause more work for employees, rather than the fabled increased efficiency and transformative gains.
​
Earlier in 2024, Amazon Web Services (AWS)​ conducted a five-week Proof-of-Concept (PoC) for ASIC, the corporate regulator for Australia.
​
Note that the focus was explicitly on measuring the quality of the generated output, rather than performance.
​
The trial involved the open source Llama2-70B model from Meta (selected as the most promising after testing), ingesting five submissions from a parliamentary inquiry into audit and consultancy firms.
​
The model was then prompted to summarise the submissions with a focus on ASIC mentions, recommendations, references to more regulation, and to include the page references and context.
​
Ten ASIC staff, of varying levels of seniority, were also given the same task.
​
Then, a group of reviewers (unaware that this exercise involved AI) blindly assessed the summaries produced by both humans and AI for coherency, length, ASIC references, regulation references and for identifying recommendations.
​
These reviewers overwhelmingly found that the human summaries trounced their AI competitors on every criteria and on every submission, scoring an 81% on an internal rubric compared with the machine’s 47%.
​
Feedback on the AI submissions included: “It was wordy and pointless - just repeating what was in the submission”, “Included analysis which did not come from the document and does not serve the purpose”, and “Made strange choices about what to highlight”.
​
The reviewers agreed that AI summaries were likely to make more work for bureaucrats (requiring fact-checking, for example), not less.
​
Thus the trial demonstrated that a human’s ability to parse and critically analyse information is unparalleled by AI...at this time.
​
​N.B.: the timeframe allowed to optimise the model was limited to one week, plus the PoC’s point-in-time results relate to the use of certain prompts using a specific LLM and selected for one specific use-case.
​
All is not lost, of course.
​
AI algorithms and LLMs are getting better all the time, every iterative release provides more optimisation and enhanced capabilities.
​
The regulator even states "ASIC is exploring and refining how artificial intelligence (AI) could be adopted into our work, recognising that AI continues to develop at a rapid pace".
​
They key takeaway here is that AI is not a panacea.

Power for the people, power to the people!

​

We need to talk about about Renewable Energy Credits (RECs).
​
RECs primarily target Scope 2 emissions, which arise from the consumption of purchased electricity.
​
When a renewable energy facility produces electricity, RECs are issued to confirm that clean energy has been generated.
​
By purchasing RECs, businesses can essentially claim ownership of the environmental benefits associated with the generated clean energy.
​
RECs currently have to be on the same continent at the same time of day.
​
Technology uses an enormous amount of power, so it generates a gigantic amount of CO2.
​
Large techs are already the largest buyers of RECs.
​
In 2023, Meta listed just 273 tonnes of "net" CO2 and claimed it had hit "net zero" - but it actually generated 3.9 million tonnes.
​
Big Tech firms are proposing a free system with no geographical limits, meaning they could offset electricity from coal power in one continent with solar power from another.
​
RECs provide mitigation in terms of meeting targets, but don’t address the core issue of overall CO2: there is no drop in real-world emissions.
​
Greenwashing bullseye.
​
In H1 2024, power generation capacity in the USA increased by the most since 2003 (and were 21% higher than the same period for 2023).
​
Though 30% of the world's electricity came from renewable sources in 2023 according to Our World in Data​, demand outpaces supply.
​
As the demand for digital and compute continues to rise, the need for Data Centre (DC) capacity and associated infrastructure will grow.
​
The precipitous proliferation of Artificial Intelligence (AI) has seen it become a real energy hog.
​
Generative AI models - text, image, speech, or video - essentially create stuff from scratch, meaning they use a lot of power at the DC.
​
Take Singapore: the second-largest DC market in Southeast Asia and the sixth-largest in Asia-Pacific.
​
Today DCs contribute to 82% of Singapore’s ICT sector emissions, and account for 7% of Singapore’s total electricity consumption.
​
Worldwide, DCs could be using a total of 1,000 terawatts hours annually by 2026.
​
This demand is roughly equivalent to the entire electricity consumption of Japan (population 125 million people), according to the International Energy Agency (IEA)​.
​
DCs have certainly got more energy efficient over time, and today underpin life as we know it.
​
Perhaps its now time for digital services to have clear energy efficiency ratings - in the same way that modern commodity appliances do - thus giving users the power (pun intended) to make informed decisions.
​
Have you considered the direct energy - or indirect environmental - impact of your usage of AI, Cloud, and other digital services?