Hydrogen power and hydrogen fuel cells have been around for decades. Hydrogen power fuel cells as an invention were invented in the 19th century. The modern hydrogen fuel cell was refined before the second world war and have been used in NASA’s space programme since Project Apollo. The space programme’s use of hydrogen power inspired General Motors to create a hydrogen fuel cell powered van in 1966. By the late 1980s, BMW had developed a hydrogen-powered engine which it trialled in its 7-series vehicles a decade later.
By the mid-2010s there were four hydrogen power passenger cars using fuel cells: Honda Clarity, Toyota Mirai, Hyundai ix35 FCEV, and the Hyundai Nexo. BMW is collaborating with Toyota to launch another four models next year.
In addition in commercial vehicles and heavy plant Hyundai, Cummins and JCB have hydrogen power offerings. JCB and Cummins have focused on internal combustion engines, while Hyundai went with hydrogen fuel cells. The aviation industry has been looking to hydrogen power via jet turbines.
Hydrogen power offers greater energy density and lower weight than batteries. Unlike batteries or power lines, hydrogen can be transported over longer distances via tanker. So someplace like Ireland with wind and tidal power potential could become an energy exporter.
The key hydrogen power problem has been investment in infrastructure and an over-reliance on batteries. Batteries bring their own set of problems and a global strategic dependency on China.
Toyota is now warning that if there isn’t imminent international investment, that China will also dominate the supply chain, exports and energy generation in the hydrogen economy as well. It feels like me reaching a historic point of no return.
Skincare you can wear: China’s sunwear boom | Jing Daily – A jacket with a wide-brim hood and built-in face shield. Leggings infused with hyaluronic acid to hydrate while shielding skin from the sun. Face masks with chin-to-temple coverage. Ice-cooling gloves designed to drop skin temperature. In China, UV protection apparel isn’t just functional — it’s fashionable, dermatological, and high-tech. Once a niche category for hikers or extreme sports enthusiasts, China’s sunwear market has exploded into a $13 billion category blending climate adaptation, anti-aging culture, and the outdoor lifestyle wave. While other apparel segments slow, the sunwear sector is projected to reach nearly 95.8 billion RMB ($13.5 billion) by 2026 expanding at a CAGR of 9%, according to iResearch.
MACAU DAILY TIMES 澳門每日時報Beware of Li Kashing’s supersized value trap – But as the initial excitement starts to fade, investors are growing nervous, wary of a billionaire family that has a poor track record on shareholders’ returns. The Li clan takes pride in the myriad of businesses and markets it operates in. But what kind of value-add can a diversified conglomerate offer when globalization is out of favor and geopolitical risks are on the rise? CK’s de-rating has accelerated since Trump’s first term, with the stock trading at just 35% of its book value even after the recent share bump. The complex business dealings have made enterprise valuation an impossible task. In a sign of deep capital market skepticism, CK seems to have trouble monetizing its assets. Its health and beauty subsidiary A.S. Watson is still privately-held, a decade after postponing an ambitious $6 billion dual listing in Hong Kong and London. Softer consumer sentiment in China, once a growth market, has become a drag. Last summer, CK Infrastructure did a secondary listing in the UK, hoping to widen its investor base. – Rare direct criticism of CK Hutchison’s conglomerate discount.
Ghost in the Shell’s Creator Wants to Revisit the Anime, But There’s One Problem – Production I.G’s CEO Mitsuhisa Ishikawa—who produced both Ghost in the Shell films—spoke at the event. Ishikawa revealed a key obstacle preventing a third film: finances. He explained that Innocence had an enormous budget, estimated at around 2 billion yen (approximately $13 million), with profits reaching a similar figure. However, the film was planned with a ten-year financial recovery period, and even after 20 years, it has yet to break even.
‘Gucci’s 25% sales collapse should shock no one’ | Jing Daily – “Gucci is so boring now.” “They’ve lost all their confidence.” “It feels desperate — just influencers and celebrities.” Comparing Gucci’s bold, visionary eras under Tom Ford and Alessandro Michele and today’s safe, uninspired iteration reveals a stark contrast. That classroom moment reflected a broader truth: Gucci’s Q1 2025 is not a temporary dip. It’s the result of a deeper structural identity crisis — arguably one of the worst brand resets in recent luxury history.
My thinking on the concept of intelligence per watt started as bullets in my notebook. It was more of a timeline than anything else at first and provided a framework of sorts from which I could explore the concept of efficiency in terms of intelligence per watt.
TL;DR (too long, didn’t read)
Our path to the current state of ‘artificial intelligence’ (AI) has been shaped by the interplay and developments of telecommunications, wireless communications, materials science, manufacturing processes, mathematics, information theory and software engineering.
Progress in one area spurred advances in others, creating a feedback loop that propelled innovation.
Over time, new use cases have become more personal and portable – necessitating a focus on intelligence per watt as a key parameter. Energy consumption directly affects industrial design and end-user benefits. Small low-power integrated circuits (ICs) facilitated fuzzy logic in portable consumer electronics like cameras and portable CD players. Low power ICs and power management techniques also helped feature phones evolve into smartphones.
A second-order effect of optimising for intelligence per watt is reducing power consumption across multiple applications. This spurs yet more new use cases in a virtuous innovation circle. This continues until the laws of physics impose limits.
Energy storage density and consumption are fundamental constraints, driving the need for a focus on intelligence per watt.
As intelligence per watt improves, there will be a point at which the question isn’t just what AI can do, but what should be done with AI? And where should it be processed? Trust becomes less about emotional reassurance and more about operational discipline. Just because it can handle a task doesn’t mean it should – particularly in cases where data sensitivity, latency, or transparency to humans is non-negotiable. A highly capable, off-device AI might be a fine at drafting everyday emails, but a questionable choice for handling your online banking.
Good ‘operational security’ outweighs trust. The design of AI systems must therefore account not just for energy efficiency, but user utility and deployment context. The cost of misplaced trust is asymmetric and potentially irreversible.
Ironically the force multiplier in intelligence per watt is people and their use of ‘artificial intelligence’ as a tool or ‘co-pilot’. It promises to be an extension of the earlier memetic concept of a ‘bicycle for the mind’ that helped inspire early developments in the personal computer industry. The upside of an intelligence per watt focus is more personal, trusted services designed for everyday use.
While not a computer, but instead to integrate several radio parts in one glass envelope vacuum valve. This had three triodes (early electronic amplifiers), two capacitors and four resistors. Inside the valve the extra resistor and capacitor components went inside their own glass tubes. Normally each triode would be inside its own vacuum valve. At the time, German radio tax laws were based on the number of valve sockets in a device, making this integration financially advantageous.
Post-war scientific boom
Between 1949 and 1957 engineers and scientists from the UK, Germany, Japan and the US proposed what we’d think of as the integrated circuit (IC). These ideas were made possible when breakthroughs in manufacturing happened. Shockley Semiconductor built on work by Bell Labs and Sprague Electric Company to connect different types of components on the one piece of silicon to create the IC.
Credit is often given to Jack Kilby of Texas Instruments as the inventor of the integrated circuit. But that depends how you define IC, with what is now called a monolithic IC being considered a ‘true’ one. Kilby’s version wasn’t a true monolithic IC. As with most inventions it is usually the child of several interconnected ideas that coalesce over a given part in time. In the case of ICs, it was happening in the midst of materials and technology developments including data storage and computational solutions such as the idea of virtual memory through to the first solar cells.
Kirby’s ICs went into an Air Force computer[ii] and an onboard guidance system for the Minuteman missile. He went on to help invent the first handheld calculator and thermal printer, both of which took advantage of progress in IC design to change our modern way of life[iii].
TTL (transistor-to-transistor logic) circuitry was invented at TRW in 1961, they licensed it out for use in data processing and communications – propelling the development of modern computing. TTL circuits powered mainframes. Mainframes were housed in specialised temperature and humidity-controlled rooms and owned by large corporates and governments. Modern banking and payments systems rely on the mainframe as a concept.
AI’s early steps
What we now thing of as AI had been considered theoretically for as long as computers could be programmed. As semiconductors developed, a parallel track opened up to move AI beyond being a theoretical possibility. A pivotal moment was a workshop was held in 1956 at Dartmouth College. The workshop focused on a hypothesis ‘every aspect of learning or any other feature of intelligence can be so precisely described that a machine can be made to simulate it’. Later on, that year a meeting at MIT (Massachusetts Institute of Technology) brought together psychologists and linguists to discuss the possibility of simulating cognitive processes using a computer. This is the origin of what we’d now call cognitive science.
Out of the cognitive approach came some early successes in the move towards artificial intelligence[iv]. A number of approaches were taken based on what is now called symbolic or classical AI:
Reasoning as search – essentially step-wise trial and error approach to problem solving that was compared to wandering through a maze and back-tracking if a dead end was found.
Natural language – where related phrases existed within a structured network.
Micro-worlds – solving for artificially simple situations, similar to economic models relying on the concept of the rational consumer.
Single layer neural networks – to do rudimentary image recognition.
By the time the early 1970s came around AI researchers ran into a number of problems, some of which still plague the field to this day:
Symbolic AI wasn’t fit for purpose solving many real-world tasks like crossing a crowded room.
Trying to capture imprecise concepts with precise language.
Commonsense knowledge was vast and difficult to encode.
Intractability – many problems require an exponential amount of computing time.
Limited computing power available – there was insufficient intelligence per watt available for all but the simplest problems.
By 1966, US and UK funding bodies were frustrated with the lack of progress on the research undertaken. The axe fell first on a project to use computers on language translation. Around the time of the OPEC oil crisis, funding to major centres researching AI was reduced by both the US and UK governments respectively. Despite the reduction of funding to the major centres, work continued elsewhere.
Mini-computers and pocket calculators
ICs allowed for mini-computers due to the increase in computing power per watt. As important as the relative computing power, ICs made mini-computers more robust, easier to manufacture and maintain. DEC (Digital Equipment Corporation) launched the first minicomputer, the PDP-8 in 1964. The cost of mini-computers allowed them to run manufacturing processes, control telephone network switching and control labouratory equipment. Mini-computers expanded computer access in academia facilitating more work in artificial life and what we’d think of as early artificial intelligence. This shift laid the groundwork for intelligence per watt as a guiding principle.
A second development helped drive mass production of ICs – the pocket calculator, originally invented at Texas Instruments. It demonstrated how ICs could dramatically improve efficiency in compact, low-power devices.
LISP machines and PCs
AI researchers required more computational power than mini-computers could provide, leading to the development of LISP machines—specialised workstations designed for AI applications. Despite improvements in intelligence per watt enabled by Moore’s Law, their specialised nature meant that they were expensive. AI researchers continued with these machines until personal computers (PCs) progressed to a point that they could run LISP quicker than LISP machines themselves. The continuous improvements in data storage, memory and processing that enabled LISP machines, continued on and surpassed them as the cost of computing dropped due to mass production.
The rise of LISP machines and their decline was not only due to Moore’s Law in effect, but also that of Makimoto’s Wave. While Gordon Moore outlined an observation that the number of transistors on a given area of silicon doubled every two years or so. Tsugio Makimoto originally observed 10-year pivots from standardised semiconductor processors to customised processors[v]. The rise of personal computing drove a pivot towards standardised architectures.
PCs and workstations extended computing beyond computer rooms and labouratories to offices and production lines. During the late 1970s and 1980s standardised processor designs like the Zilog Z80, MOS Technology 6502 and the Motorola 68000 series drove home and business computing alongside Intel’s X86 processors.
Personal computing started in businesses when office workers brought a computer to use early computer programmes like the VisiCalc spreadsheet application. This allowed them to take a leap forward in not only tabulating data, but also seeing how changes to the business might affect financial performance.
Businesses then started to invest more in PCs for a wide range of uses. PCs could emulate the computer terminal of a mainframe or minicomputer, but also run applications of their own.
Typewriters were being placed by word processors that allowed the operator to edit a document in real time without resorting to using correction fluid.
A Bicycle for the Mind
Steve Jobs at Apple was as famous for being a storyteller as he was for being a technologist in the broadest sense. Internally with the Mac team he shared stories and memetic concepts to get his ideas across in everything from briefing product teams to press interviews. As a concept, a 1990 filmed interview with Steve Jobs articulates the context of this saying particularly well.
In reality, Jobs had been telling the story for a long time through the development of the Apple II and right from the beginning of the Mac. There is a version of the talk that was recorded some time in 1980 when the personal computer was still a very new idea – the video was provided to the Computer History Museum by Regis McKenna[vi].
The ‘bicycle for the mind’ concept was repeated in early Apple advertisements for the time[vii] and even informed the Macintosh project codename[viii].
Jobs articulated a few key concepts.
Buying a computer creates, rather than reduces problems. You needed software to start solving problems and making computing accessible. Back in 1980, you programmed a computer if you bought one. Which was the reason why early personal computer owners in the UK went on to birth a thriving games software industry including the likes of Codemasters[ix]. Done well, there should be no seem in the experience between hardware and software.
The idea of a personal, individual computing device (rather than a shared resource). My own computer builds on my years of how I have grown to adapt and use my Macs, from my first sit-up and beg Macintosh, to the MacBook Pro that I am writing this post on. This is even more true most people and their use of the smartphone. I am of an age, where my iPhone is still an appendage and emissary of my Mac. My Mac is still my primary creative tool. A personal computer is more powerful than a shared computer in terms of the real difference made.
At the time Jobs originally did the speech, PCs were underpowered for anything but data processing (through spreadsheets and basic word processor applications). But that didn’t stop his idea for something greater.
Jobs idea of the computer as an adjunct to the human intellect and imagination still holds true, but it doesn’t neatly fit into the intelligence per watt paradigm. It is harder to measure the effort developing prompts, or that expended evaluating, refining and filtering generative AI results. Of course, Steve Jobs Apple owed a lot to the vision shown in Doug Engelbart’s ‘Mother of All Demos’[x].
Networks
Work took a leap forward with office networked computers pioneered by Macintosh office by Apple[xi]. This was soon overtaken by competitors. This facilitated work flow within an office and its impact can still be seen in offices today, even as components from print management to file storage have moved to cloud-based services.
At the same time, what we might think of as mobile was starting to gain momentum. Bell Labs and Motorola came up with much of the technology to create cellular communications. Martin Cooper of Motorola made the first phone call on a cellular phone to a rival researcher at Bell Labs. But Motorola didn’t sell the phone commercially until 1983, as a US-only product called the DynaTAC 8000x[xii]. This was four years after Japanese telecoms company NTT launched their first cellular network for car phones. Commercial cellular networks were running in Scandinavia by 1981[xiii].
In the same way that the networked office radically changed white collar work, the cellular network did a similar thing for self-employed plumbers, electricians and photocopy repair men to travelling sales people. If they were technologically advanced, they may have had an answer machine, but it would likely have to be checked manually by playing back the tape.
Often it was a receptionist in their office if they had one. Or more likely, someone back home who took messages. The cell phone freed homemakers in a lot of self-employed households to go out into the workplace and helped raise household incomes.
Fuzzy logic
The first mainstream AI applications emerged from fuzzy logic, introduced by Lofti A. Zadeh in 1965 mathematical paper. Initial uses were for industrial controls in cement kilns and steel production[xiv]. The first prominent product to rely on fuzzy logic was the Zojirushi Micom Electric Rice Cooker (1983), which adjusted cooking time dynamically to ensure perfect rice.
Fuzzy logic reacted to changing conditions in a similar way to people. Through the 1980s and well into the 1990s, the power of fuzzy logic was under appreciated outside of Japanese product development teams. In a quote a spokesperson for the American Electronics Association’s Tokyo office said to the Washington Post[xv].
“Some of the fuzzy concepts may be valid in the U.S.,”
“The idea of better energy efficiency, or more precise heating and cooling, can be successful in the American market,”
“But I don’t think most Americans want a vacuum cleaner that talks to you and says, ‘Hey, I sense that my dust bag will be full before we finish this room.’ “
The end of the 1990s, fuzzy logic was embedded in various consumer devices:
Air-conditioner units – understands the room, the temperature difference inside-and-out, humidity. It then switches on-and-off to balance cooling and energy efficiency.
CD players – enhanced error correction on playback dealing with imperfections on the disc surface.
Dishwashers – understood how many dishes were loaded, their type of dirt and then adjusts the wash programme.
Toasters – recognised different bread types, the preferable degree of toasting and performs accordingly.
TV sets – adjust the screen brightness to the ambient light of the room and the sound volume to how far away the viewer is sitting from the TV set.
Vacuum cleaners – vacuum power that is adjusted as it moves from carpeted to hard floors.
Video cameras – compensate for the movement of the camera to reduce blurred images.
Fuzzy logic sold on the benefits and concealed the technology from western consumers. Fuzzy logic embedded intelligence in the devices. Because it worked on relatively simple dedicated purposes it could rely on small lower power specialist chips[xvi] offering a reasonable amount of intelligence per watt, some three decades before generative AI. By the late 1990s, kitchen appliances like rice cookers and microwave ovens reached ‘peak intelligence’ for what they needed to do, based on the power of fuzzy logic[xvii].
Fuzzy logic also helped in business automation. It helped to automatically read hand-written numbers on cheques in banking systems and the postcodes on letters and parcels for the Royal Mail.
Decision support systems & AI in business
Decision support systems or Business Information Systems were being used in large corporates by the early 1990s. The techniques used were varied but some used rules-based systems. These were used in at least some capacity to reduce manual office work tasks. For instance, credit card approvals were processed based on rules that included various factors including credit scores. Only some credit card providers had an analyst manually review the decision made by system. However, setting up each use case took a lot of effort involving highly-paid consultants and expensive software tools. Even then, vendors of business information systems such as Autonomy struggled with a high rate of projects that failed to deliver anything like the benefits promised.
Three decades on, IBM had a similar problem with its Watson offerings, with particularly high-profile failure in mission-critical healthcare applications[xviii]. Secondly, a lot of tasks were ad-hoc in nature, or might require transposing across disparate separate systems.
The rise of the web
The web changed everything. The underlying technology allowed for dynamic data.
Software agents
Examples of intelligence within the network included early software agents. A good example of this was PapriCom. PapriCom had a client on the user’s computer. The software client monitored price changes for products that the customer was interested in buying. The app then notified the user when the monitored price reached a price determined by the customer. The company became known as DealTime in the US and UK, or Evenbetter.com in Germany[xix].
The PapriCom client app was part of a wider set of technologies known as ‘push technology’ which brought content that the netizen would want directly to their computer. In a similar way to mobile app notifications now.
Web search
The wealth of information quickly outstripped netizen’s ability to explore the content. Search engines became essential for navigating the new online world. Progress was made in clustering vast amounts of cheap Linux powered computers together and sharing the workload to power web search amongst them. As search started to trying and make sense of an exponentially growing web, machine learning became part of the developer tool box.
Researchers at Carnegie-Mellon looked at using games to help teach machine learning algorithms based on human responses that provided rich metadata about the given item[xx]. This became known as the ESP game. In the early 2000s, Yahoo! turned to web 2.0 start-ups that used user-generated labels called tags[xxi] to help organise their data. Yahoo! bought Flickr[xxii] and deli.ico.us[xxiii].
All the major search engines looked at how deep learning could help improve search results relevance.
Given that the business model for web search was an advertising-based model, reducing the cost per search, while maintaining search quality was key to Google’s success. Early on Google focused on energy consumption, with its (search) data centres becoming carbon neutral in 2007[xxiv]. This was achieved by a whole-system effort: carefully managing power management in the silicon, storage, networking equipment and air conditioning to maximise for intelligence per watt. All of which were made using optimised versions of open-source software and cheap general purpose PC components ganged together in racks and operating together in clusters.
General purpose ICs for personal computers and consumer electronics allowed easy access relatively low power computing. Much of this was down to process improvements that were being made at the time. You needed the volume of chips to drive innovation in mass-production at a chip foundry. While application-specific chips had their uses, commodity mass-volume products for uses for everything from embedded applications to early mobile / portable devices and computers drove progress in improving intelligence-per-watt.
Makimoto’s tsunami back to specialised ICs
When I talked about the decline of LISP machines, I mentioned the move towards standardised IC design predicted by Tsugio Makimoto. This led to a surge in IC production, alongside other components including flash and RAM memory. From the mid-1990s to about 2010, Makimoto’s predicted phase was stuck in ‘standardisation’. It just worked. But several factors drove the swing back to specialised ICs.
Lithography processes got harder: standardisation got its performance and intelligence per watt bump because there had been a steady step change in improvements in foundry lithography processes that allowed components to be made at ever-smaller dimensions. The dimensions are a function wavelength of light used. The semiconductor hit an impasse when it needed to move to EUV (extreme ultra violet) light sources. From the early 1990s on US government research projects championed development of key technologies that allow EUV photolithography[xxv]. During this time Japanese equipment vendors Nikon and Canon gave up on EUV. Sole US vendor SVG (Silicon Valley Group) was acquired by ASML, giving the Dutch company a global monopoly on cutting edge lithography equipment[xxvi]. ASML became the US Department of Energy research partner on EUV photo-lithography development[xxvii]. ASML spent over two decades trying to get EUV to work. Once they had it in client foundries further time was needed to get commercial levels of production up and running. All of which meant that production processes to improve IC intelligence per watt slowed down and IC manufacturers had to start about systems in a more holistic manner. As foundry development became harder, there was a rise in fabless chip businesses. Alongside the fabless firms, there were fewer foundries: Global Foundries, Samsung and TSMC (Taiwan Semiconductor Manufacturing Company Limited). TSMC is the worlds largest ‘pure-play’ foundry making ICs for companies including AMD, Apple, Nvidia and Qualcomm.
Progress in EDA (electronic design automation). Production process improvements in IC manufacture allowed for an explosion in device complexity as the number of components on a given size of IC doubled every 18 months or so. In the mid-to-late 1970s this led to technologists thinking about the idea of very large-scale integration (VLSI) within IC designs[xxviii]. Through the 1980s, commercial EDA software businesses were formed. The EDA market grew because it facilitated the continual scaling of semiconductor technology[xxix]. Secondly, it facilitated new business models. Businesses like ARM Semiconductor and LSI Logic allowed their customers to build their own processors based on ‘blocs’ of proprietary designs like ARM’s cores. That allowed companies like Apple to focus on optimisation in their customer silicon and integration with software to help improve the intelligence per watt[xxx].
Increased focus on portable devices. A combination of digital networks, wireless connectivity, the web as a communications platform with universal standards, flat screen displays and improving battery technology led the way in moving towards more portable technologies. From personal digital assistants, MP3 players and smartphone, to laptop and tablet computers – disconnected mobile computing was the clear direction of travel. Cell phones offered days of battery life; the Palm Pilot PDA had a battery life allowing for couple of days of continuous use[xxxi]. In reality it would do a month or so of work. Laptops at the time could do half a day’s work when disconnected from a power supply. Manufacturers like Dell and HP provided spare batteries for travellers. Given changing behaviours Apple wanted laptops that were easy to carry and could last most of a day without a charge. This was partly driven by a move to a cleaner product design that wanted to move away from swapping batteries. In 2005, Apple moved from PowerPC to Intel processors. During the announcement at the company’s worldwide developer conference (WWDC), Steve Jobs talked about the focus on computing power per watt moving forwards[xxxii].
Apple’s first in-house designed IC, the A4 processor was launched in 2010 and marked the pivot of Makimoto’s wave back to specialised processor design[xxxiii]. This marked a point of inflection in the growth of smartphones and specialised computing ICs[xxxiv].
New devices also meant new use cases that melded data on the web, on device, and in the real world. I started to see this in action working at Yahoo! with location data integrated on to photos and social data like Yahoo! Research’s ZoneTag and Flickr. I had been the Yahoo! Europe marketing contact on adding Flickr support to Nokia N-series ‘multimedia computers’ (what we’d now call smartphones), starting with the Nokia N73[xxxv]. A year later the Nokia N95 was the first smartphone released with a built-in GPS receiver. William Gibson’s speculative fiction story Spook Country came out in 2007 and integrated locative art as a concept in the story[xxxvi].
Real-world QRcodes helped connect online services with the real world, such as mobile payments or reading content online like a restaurant menu or a property listing[xxxvii].
I labelled the web-world integration as a ‘web-of-no-web’[xxxviii] when I presented on it back in 2008 as part of an interactive media module, I taught to an executive MBA class at Universitat Ramon Llull in Barcelona[xxxix]. In China, wireless payment ideas would come to be labelled O2O (offline to online) and Kevin Kelly articulated a future vision for this fusion which he called Mirrorworld[xl].
Deep learning boom
Even as there was a post-LISP machine dip in funding of AI research, work on deep (multi-layered) neural networks continued through the 1980s. Other areas were explored in academia during the 1990s and early 2000s due to the large amount of computing power needed. Internet companies like Google gained experience in large clustered computing, AND, had a real need to explore deep learning. Use cases include image recognition to improve search and dynamically altered journeys to improve mapping and local search offerings. Deep learning is probabilistic in nature, which dovetailed nicely with prior work Microsoft Research had been doing since the 1980s on Bayesian approaches to problem-solving[xli].
A key factor in deep learning’s adoption was having access to powerful enough GPUs to handle the neural network compute[xlii]. This has allowed various vendors to build Large Language Models (LLMs). The perceived strategic importance of artificial intelligence has meant that considerations on intelligence per watt has become a tertiary consideration at best. Microsoft has shown interest in growing data centres with less thought has been given on the electrical infrastructure required[xliii].
Google’s conference paper on attention mechanisms[xliv] highlighted the development of the transformer model. As an architecture it got around problems in previous approaches, but is computationally intensive. Even before the paper was published, the Google transformer model had created fictional Wikipedia entries[xlv]. A year later OpenAI built on Google’s work with the generative pre-trained transformer model better known as GPT[xlvi].
Since 2018 we’ve seen successive GPT-based models from Amazon, Anthropic, Google, Meta, Alibaba, Tencent, Manus and DeepSeek. All of these models were trained on vast amounts of information sources. One of the key limitations for building better models was access to training material, which is why Meta used pirated copies of e-books obtained using bit-torrent[xlvii].
These models were so computationally intensive that the large-scale cloud service providers (CSPs) offering these generative AI services were looking at nuclear power access for their data centres[xlviii].
The current direction of development in generative AI services is raw computing power, rather than having a more energy efficient focus of intelligence per watt.
Technology consultancy / analyst Omdia estimated how many GPUs were bought by hyperscalers in 2024[xlix].
Company
Number of Nvidia GPUs bought
Number of AMD GPUs bought
Number of self-designed custom processing chips bought
Amazon
196,000
–
1,300,000
Alphabet (Google)
169,000
–
1,500,000
ByteDance
230,000
–
–
Meta
224,000
173,000
1,500,000
Microsoft
485,000
96,000
200,000
Tencent
230,000
–
–
These numbers provide an indication of the massive deployment on GPT-specific computing power. Despite the massive amount of computing power available, services still weren’t able to cope[l] mirroring some of the service problems experienced by early web users[li] and the Twitter ‘whale FAIL’[lii] phenomenon of the mid-2000s. The race to bigger, more powerful models is likely to continue for the foreseeable future[liii].
There is a second class of players typified by Chinese companies DeepSeek[liv] and Manus[lv] that look to optimise the use of older GPT models to squeeze the most utility out of them in a more efficient manner. Both of these services still rely on large cloud computing facilities to answer queries and perform tasks.
Agentic AI
Thinking on software agents went back to work being done in computer science in the mid-1970s[lvi]. Apple articulated a view[lvii]of a future system dubbed the ‘Knowledge Navigator’[lviii] in 1987 which hinted at autonomous software agents. What we’d now think of as agentic AI was discussed as a concept at least as far back as 1995[lix], this was mirrored in research labs around the world and was captured in a 1997 survey of research on intelligent software agents was published[lx]. These agents went beyond the vision that PapriCom implemented.
A classic example of this was Wildfire Communications, Inc. who created a voice enabled virtual personal assistant in 1994[lxi]. Wildfire as a service was eventually shut down in 2005 due to an apparent decline in subscribers using the service[lxii]. In terms of capability, Wildfire could do tasks that are currently beyond Apple’s Siri. Wildfire did have limitations due to it being an off-device service that used a phone call rather than an internet connection, which limited its use to Orange mobile service subscribers using early digital cellular mobile networks.
Almost a quarter century later we’re now seeing devices that are looking to go beyond Wildfire with varying degrees of success. For instance, the Rabbit R1 could order an Uber ride or groceries from DoorDash[lxiii]. Google Duplex tries to call restaurants on your behalf to make reservations[lxiv] and Amazon claims that it can shop across other websites on your behalf[lxv]. At the more extreme end is Boeing’s MQ-28[lxvi] and the Loyal Wingman programme[lxvii]. The MQ-28 is an autonomous drone that would accompany US combat aircraft into battle, once it’s been directed to follow a course of action by its human colleague in another plane.
The MQ-28 will likely operate in an electronic environment that could be jammed. Even if it wasn’t jammed the length of time taken to beam AI instructions to the aircraft would negatively impact aircraft performance. So, it is likely to have a large amount of on-board computing power. As with any aircraft, the size of computing resources and their power is a trade-off with the amount of fuel or payload it will carry. So, efficiency in terms of intelligence per watt becomes important to develop the smallest, lightest autonomous pilot.
As well as a more hostile world, we also exist in a more vulnerable time in terms of cyber security and privacy. It makes sense to have critical, more private AI tasks run on a local machine. At the moment models like DeepSeek can run natively on a top-of-the-range Mac workstation with enough memory[lxviii].
This is still a long way from the vision of completely local execution of ‘agentic AI’ on a mobile device because the intelligence per watt hasn’t scaled down to that level to useful given the vast amount of possible uses that would be asked of the Agentic AI model.
Maximising intelligence per watt
There are three broad approaches to maximise the intelligence per watt of an AI model.
Take advantage of the technium. The technium is an idea popularised by author Kevin Kelly[lxix]. Kelly argues that technology moves forward inexorably, each development building on the last. Current LLMs such as ChatGPT and Google Gemini take advantage of the ongoing technium in hardware development including high-speed computer memory and high-performance graphics processing units (GPU). They have been building large data centres to run their models in. They build on past developments in distributed computing going all the way back to the 1962[lxx].
Optimise models to squeeze the most performance out of them. The approach taken by some of the Chinese models has been to optimise the technology just behind the leading-edge work done by the likes of Google, OpenAI and Anthropic. The optimisation may use both LLMs[lxxi] and quantum computing[lxxii] – I don’t know about the veracity of either claim.
Specialised models. Developing models by use case can reduce the size of the model and improve the applied intelligence per watt. Classic examples of this would be fuzzy logic used for the past four decades in consumer electronics to Mistral AI[lxxiii] and Anduril’s Copperhead underwater drone family[lxxiv].
Even if an AI model can do something, should the model be asked to do so?
We have a clear direction of travel over the decades to more powerful, portable computing devices –which could function as an extension of their user once intelligence per watt allows it to be run locally.
Having an AI run on a cloud service makes sense where you are on a robust internet connection, such as using the wi-fi network at home. This makes sense for general everyday task with no information risk, for instance helping you complete a newspaper crossword if there is an answer you are stuck on and the intellectual struggle has gone nowhere.
A private cloud AI service would make sense when working, accessing or processing data held on the service. Examples of this would be Google’s Vertex AI offering[lxxv].
On-device AI models make sense in working with one’s personal private details such as family photographs, health information or accessing apps within your device. Apps like Strava which share data, have been shown to have privacy[lxxvi] and security[lxxvii] implications. ***I am using Strava as an example because it is popular and widely-known, not because it is a bad app per se.***
While businesses have the capability and resources to have a multi-layered security infrastructure to protect their data most[lxxviii]of[lxxix] the[lxxx] time[lxxxi], individuals don’t have the same security. As I write this there are privacy concerns[lxxxii] expressed about Waymo’s autonomous taxis. However, their mobile device is rarely out of physical reach and for many their laptop or tablet is similarly close. All of these devices tend to be used in concert with each other. So, for consumers having an on-device AI model makes the most sense. All of which results in a problem, how do technologists squeeze down their most complex models inside a laptop, tablet or smartphone?
Clutch Cargo was an animated series first broadcast on American television in 1959. Clutch Cargo was created by Cambria Productions – who were a start-up animation studio. Cambria used a number of techniques to radically reduce the cost of producing the animated series.
A key consideration was reducing the amount of movement that needed to be animated. There were some obvious visual motifs used to do this:
Characters were animated from waist height up for the majority of the films, this reduced the need to animate legs, walking or running.
Much of the movement was moving the camera around, towards or away from a static picture.
To show an explosion, they shook the camera, rather than animate the concussive effect of the blast.
Fire wasn’t animated, instead smoke would be put in front of the camera. Fake snow was sprinkled so that bad weather didn’t need to be drawn.
Cameraman Ted Gillette came up with the idea of Syncro-Vox. The voice actors head would be held steady, they would have a vivid lipstick applied and then say their lines. Gillette then put their mouths on top of the animated figures. Cambria made use of it in all their animations with the exception of The New Three Stooges – an animated series that allowed Moe Howard, Larry Fine and Joe DeRita to be voice actors after their movie contracts finished and they were affected by ill health.
These choices meant that Clutch Cargo cost about 10 per cent of what it would have cost Disney to animate. The visual hacks to cut costs were also helped in the way the scripts were developed. Clutch Cargo avoided doing comedy, instead focusing on Tin-Tin-like adventures. ‘Physical’ comedy gags create a lot of movement to animate. By focusing on the storytelling of Clutch Cargo. The young audience weren’t bothered by the limited animation, as they were captivated into suspending their beliefs.
Ozempic Could Crush the Junk Food Industry. But It Is Fighting Back. – The New York Times – Lars Fruergaard Jorgensen, the chief executive of Novo Nordisk, which makes Ozempic and Wegovy, told Bloomberg that food-industry executives had been calling him. “They are scared about it,” he said. Around the same time, Walmart’s chief executive in the United States, John Furner, said that customers on GLP-1s were putting less food into their carts. Sales are down in sweet baked goods and snacks, and the industry is weathering a downturn. By one market-research firm’s estimate, food-and-drink innovation in 2024 reached an all-time nadir, with fewer new products coming to market than ever before.
Ozempic users like Taylor aren’t just eating less. They’re eating differently. GLP-1 drugs seem not only to shrink appetite but to rewrite people’s desires. They attack what Amy Bentley, a food historian and professor at New York University, calls the industrial palate: the set of preferences created by our acclimatization, often starting with baby food, to the tastes and textures of artificial flavors and preservatives. Patients on GLP-1 drugs have reported losing interest in ultraprocessed foods, products that are made with ingredients you wouldn’t find in an ordinary kitchen: colorings, bleaching agents, artificial sweeteners and modified starches. Some users realize that many packaged snacks they once loved now taste repugnant.
Apple resumes advertising on Elon Musk’s X after 15-month pause – 9to5Mac – the negative reaction to this that I have seen from Mac and iPhone users that I know is interesting. It’s the scales have dropped from their eyes about Apple’s performative progressive values. Yet the signs have been out there for years – in particular with regards anything that is even tangentially connected to China.
Zuckerberg’s rightward policy shift hits Meta staffers, targets Apple | CNBC – employees who might otherwise leave because of their disillusionment with policy changes are concerned about quitting now because of how they will be perceived by future employers given that Meta has said publicly that it’s weeding out “low performers.” Meta, like many of its tech peers, began downsizing in 2022 and has continued to trim around the edges. The company cut 21,000 jobs, or nearly a quarter of its workforce, in 2022 and 2023. Among those who lost their jobs were members of the civic integrity group, which was known to be outspoken in its criticism of Zuckerberg’s leadership. Some big changes are now taking place that appear to directly follow the lead of Trump at the expense of company employees and users of the platforms, the people familiar with the matter said.
Foreign workers in Singapore parlance are people who come from around Southeast Asia and South Asia to do blue collar and pink collar jobs in the city state.
In a number of Asian countries including Hong Kong and Singapore; Filipino and Indonesian workers came to care for old people at home, look after children and conduct household tasks.
This group of foreign workers freed up middle class married women in Singapore and other countries to participate more to their economy, capitalising on their education and ability to earn more in fast-growing economies. They had higher levels of workforce participation than their female counterparts in Japan and South Korea.
The Philippines relies almost five-fold more on remittances for its GDP than similar countries like Indonesia.
What’s less reflected upon is the social upheaval and challenges that these foreign workers face in their new homes. They are in a different culture, away from friends and family as a support network. They have tremendous pressure to remit as much money as possible home.
They only have each other to rely upon. This skate team is just one of the activities that foreign workers do. From informal gatherings with friends to sophisticated beauty pageants, volleyball and basketball leagues. More Singapore related content can be found here.
Beauty
China’s beauty market is a sight for sore eyes | FT – The brand keeps prices of its products, from face powders to creams, closer to those of premium international brands, in line with L’Oréal’s Lancôme and Shiseido’s Nars. The rise of a domestic premium brand points to a significant shift in mainland shoppers’ buying habits as well as highlighting improvements in the quality of domestic products
Y3K: Futuristic fashion trend sweeps China | Jing Daily – Inspired by AI, VR, and the metaverse, and propelled by K-pop idols and Korean brands, Y3K is rapidly gaining popularity among Gen Z. – very William Gibson ‘Burning Chrome’ era
Economics
Diverging demographic destinies: Cars and the middle class | WARC – According to Pew, the American middle class has shrunk significantly in the last few decades. The top 20% of earners now take more than 50% of aggregate income because theirs has grown faster. 88% of Americans have less than $2000 in their checking account and 50% have less than $500 in savings. The average cost of a new car in 1984 was $6000 and the average household income was $27k. Today average household income is $80k [Fed] but averages conceal the widened gap between maxima and minima: the median income per person is around $35k [Census]. The average price of a new car is almost $50k, which is surprising enough that CNN wrote an article about it. They explain that “much of the reason Americans are paying nearly $50k for a car is that automakers decided to go all-in on expensive cars. The more they charge for a car, the more money they make off it.”
Whereas forty years ago an average new car cost about a fifth of an average annual salary, a new car is now prohibitively expensive for most. That’s why Americans have a record $1.6 trillion of outstanding car debt and delinquencies are rising.
Biden’s Move to Block US Steel Deal Is No Way to Treat Japan – Bloomberg – In the executive order preventing the deal on spurious national security grounds, staffers for President Joe Biden appeared to accidentally copy-and-paste the title of a previous presidential order — one ordering a Chinese crypto mining company to vacate property near an Air Force base. The left the Nippon Steel directive entitled: “Regarding the acquisition of certain real property of Cheyenne leads by MineOne Cloud Computing Investment.”
Luxury
Interesting research from two sources that don’t quite square with each other. Walpole’s The State of London Luxury 2024 report came out and painted a rosy picture about the ultra high end aspect of the London property market. Meanwhile over at the FT, Why London’s property market is stagnating points at the same end of the market as being moribund in nature.
Welcome to my September 2024 newsletter, this newsletter marks my 14th issue. When I lived in Hong Kong and dealt with Chinese accounts, the number 4 was considered unlucky, rather like 13. 14 is even worse due to it sounding so similar to ‘is dead’ or ‘will be dead’ or ‘will be certain to die’ depending on the variant of the spoken language used. In other cultures the symbolism of 14 is more nebulous at best.
September got off to an odd start, we seem to have had all the seasons, rather than settling gently into the run up to autumn. I managed to avoid traditional mooncake during mid-autumn festival celebrations that I attended. My waistline was thankful for the #ROMO (relief of missing out).
New reader?
If this is the first newsletter, welcome! You can find my regular writings here and more about me here.
Things I’ve written.
Being on the ground in Merseyside as the Southport stabbings unfolded gave me a different perspective on things.
How generative AI features are affecting the Google search experience and much more.
My reading for September 2024 slacked off a bit as real world obligations kicked in.
The Old Woman With the Knife worked on a number of levels for me. Firstly, I loved its portrayal of modern Korean society, from the aging population to the Confucian view of seniority that makes everyday interactions more complex than other Asian societies. Without revealing too much, the old woman in question is someone in the twilight of her career and how she is coping with new up-and-coming rivals at work.
Panic! edited by Michael Lewis. Michael Lewis became famous when he wrote an account of his career in investment banking in Liar’s Poker. His career overlapped with the 1987 financial crash. Since then he has been a writer who has documented key turns in the economy. Because of this background Lewis was the ideal person to curate a history of financial crisis from contemporary accounts at the time. Panic! covers the 1987 financial crash, the 1998 debt crisis, the dot com bubble, and the subprime mortgage crisis of 2007/8. I read the book in short bursts mainly due to asks on my time, rather than the nature of the book. Panic! seemed pertinent to read now. The publication of Pegasus Research’s iconic quantitative research on ‘burn rates’ in March 2000 on dot.com company burn rates makes it highly relevant to revisit when we are in hype cycles such as those surrounding health tech, fintech, crypto and more – if for no other reason than pointing out the folly of trying to pick winners in hype-driven public markets with a high degree of opacity.
Klara and the Sun by Kazuo Ishiguro moved from my to be read pile to must read pile given everything that has been going on with generative AI over the the past couple of years that it has sat on my bookshelf. Ishiguro uses speculative fiction to explore the different kinds of love and attachment, alongside loss. From a machine learning perspective it poses interesting questions about applying observational learning rather than rules based learning in systems that are supposed to exist in the real world. Klara is an ‘artificial friend’ for a child who is going through ‘levelling up’. Levelling up could be seen as a euphemism for everything from the cramming schools popular in Asian education systems to the challenges humans face in an information heavy environment. Ultimately there is something more human and child-like in Klara’s experiences than the human co-protagonists.
Things I have been inspired by.
AI proficiency.
Section, the education company founded by Professor Scott Galloway has AI proficiency as a key element in its offering. They have put together research to show how low the current level of proficiency is. They consider this research a rallying cry; but the results could also be reflective of a technology adoption curve that isn’t moving at the speed of hype, which is what came through in my examination of public discussions during the summer.
Secondly, research from the Upwork Research Institute implies a higher a higher adoption rate of generative AI, but lower success rate with the outputs generating inefficiencies rather than productivity gains. Part of the problem seems to come from organisational leadership and the way generative AI is being implemented.
WARC have published a report which looks at What’s working in generative AI from a marketing perspective. Some of the ideas like synthetic data in market research are not quite in prime time yet and generative AI’s large carbon footprint can’t be ignored.
Digitalisation and brands
Harvard Business Review published research that indicated a weaker relationship between profit share and brand in certain types of businesses. On the face of it, this supports Scott Galloway’s ‘end of brand‘ hypothesis. WARC covered the research paper in depth pointing out that for each percentage gain in market share highly digitised businesses gain 0.19% increase in profit compared to 0.26% in less digital businesses. This seems to be due to a multitude of factors:
Efficiency gains due to digitalisation have an effect on the existing profit prior to the market share. Efficiency is the main selling point of much digital automation from CRM systems to performance advertising.
Market power of larger companies ( a la Google).
Perceptions of quality – digital-only companies might look more reputable due to the lack of real-world signals to the contrary
Market share (and brand) still matters, but it hits different depending on the business. B2B and growth hacking business approaches gain less than consumer orientated businesses. A larger dataset of Kantar-sourced data analysed by Oxford University researchers found that better brand effects were down to ‘difference’ as in how customers see – and experience the brand – as being different enough from competitors.
FEAST
FEAST is an occasional magazine and curated set of events all about food and its ingredients. If you are a strategist working with a food or beverage client it’s well worth exploring their archive as a source of inspiration for insights given its in-depth and thoughtful arts-based approach.
Streaming plateau
I finally got to dive into Ofcom’s Media Nations research report. I recommend that if you are involved in the advertising-media industrial complex in any way shape or form, spend some time reading it. On the plus side, survey respondents consider accurate balanced news as a key part of the public service mandate of radio and television. Secondly during 2023, broadcast brought many of us together still for key events including the first episodes of tentpole series.
Ofcom
More dispiritingly, I realised that amongst the tentpole TV series was season 23 or I’m a Celebrity, Get Me Out of Here. TV advertising revenues declined faster than online video revenues grew and subscription based video on demand take-up plateaued. It’s pure speculation on my part, but this might have been reached because COVID accelerated adoption.
Finally, as a film fan who buys Blu-Rays of films that don’t appear on streaming services consistently, the amount of time watching DVDs and and Blu-Rays have a very small usage across all age groups. I don’t think that’s good news for arthouse and cult cinema.
Things I have watched.
The Crow – less of a remake of the iconic Brandon Lee film and more of a reinvention of the franchise. It’s a good if unengaging film, many of the young adult audience I went to see it with won’t have the original or the comic books as a reference point. It has been described on other parts of the net as ‘the worst movie of the year‘. This is probably a bit unfair; there is a lot of ‘straight to Netflix’ dross out there. FKA Twigs character in the film grew on me as I watched it.
The last time I saw The Terminator in a cinema was in a double bill with the then newly launched Terminator 2 at the then new Odeon multiplex in Bromborough. I got to see it again thanks to a 4K restoration. Despite having seen The Terminator several times on tape, DVD and Blu-Ray – this time it hit different. It hit harder and it was all down the way the screen filled my vision and the punch of the sound track. Despite in-home cinema set-ups, you just can’t get at home unless you live in a large industrial unit sized home. The analogue special effects held up surprisingly well and the plot was just as taunt as I remembered it. There was less people in the cinema than for the screening than for TheCrow.
As I write this, I have just watched episode one of the latest series of Slow Horses. It has gotten off to the high standard set by the previous series and book. Time to put on my Roddy Ho t-shirt again.
Useful tools.
Table Capture
If you’ve ever tried to cut and paste a table and data from a website into an spreadsheet and then spent the rest of the afternoon parsing it in cell-by-cell you will appreciate the benefit of this browser extension.
Humaniser for GPT created content
If you’re reading this, chances are that you’ve used services like Claude, ChatGPT or Gemini as a starting point for copy, or to summarise documents. UndetectableGPT looks at providing alternatives to ‘tell-tale’ phraseology in generative AI copy.
Data analysis
Groupt will take a CSV file and categorise the data including visualisation, so you only have to focus on wrapping a narrative around it to fit into the wider storytelling of your presentation.
The sales pitch.
I am now taking bookings for strategic engagements from January 2025 onwards; or discussions on permanent roles. Contact me here.
Ok this is the end of my September 2024 newsletter, I hope to see you all back here again in a month. Be excellent to each other and onward into October and crispness of a bright autumn morning!
