Category: English

Source: The Conversation – Global Perspectives – By Milad Haghani, Associate Professor and Principal Fellow in Urban Risk and Resilience, The University of Melbourne

The Hong Kong high-rise fire, which spread across multiple buildings in a large residential complex, has killed dozens, with hundreds reported missing.

The confirmed death toll is now 44, with close to 300 people still unaccounted for and dozens in hospital with serious injuries.

This makes it one of Hong Kong’s deadliest building fires in living memory, and already the worst since the Garley Building fire in 1996.

Although more than 900 people have been reportedly evacuated from the Wang Fuk Court, it’s not clear how many residents remain trapped.

This catastrophic fire – which is thought to have spread from building to building via burning bamboo scaffolding and fanned by strong winds – highlights how difficult it is to evacuate high-rise buildings in an emergency.

When the stakes are highest

Evacuations of high-rises don’t happen every day, but occur often enough. And when they do, the consequences are almost always severe. The stakes are highest in the buildings that are full at predictable times: residential towers at night, office towers in the day.

We’ve seen this in the biggest modern examples, from the World Trade Center in the United States to Grenfell Tower in the United Kingdom.

The patterns repeat: once a fire takes hold, getting thousands of people safely down dozens of storeys becomes a race against time.

But what actually makes evacuating a high-rise building so challenging?

It isn’t just a matter of “getting people out”. It’s a collision between the physical limits of the building and the realities of human behaviour under stress.

It’s a long way down to safety

The biggest barrier is simply vertical distance. Stairwells are the only reliable escape route in most buildings.

Stair descent in real evacuations is far slower than most people expect. Under controlled or drill conditions people move down at around 0.4–0.7 metres per second. But in an actual emergency, especially in high-rise fires, this can drop sharply.

During 9/11, documented speeds at which survivors went down stairs were often slower than 0.3 m/s. These slow-downs accumulate dramatically over long vertical distances.

Fatigue is a major factor. Prolonged walking significantly reduces the speed of descent. Surveys conducted after incidents confirm that a large majority of high-rise evacuees stop at least once. During the 2010 fire of a high-rise in Shanghai, nearly half of older survivors reported slowing down significantly.

Long stairwells, landings, and the geometry of high-rise stairs all contribute to congestion, especially when flows from multiple floors merge into a single shaft.

Slower movers include older adults, people with physical or mobility issues and groups evacuating together. These reduce the overall pace of descent compared with the speeds typically assumed for able-bodied individuals. This can create bottlenecks. Slow movers are especially relevant in residential buildings, where diverse occupants mean movement speeds vary widely.

Visibility matters too. Experimental studies show that reduced lighting significantly slows down people going down stairs. This suggests that when smoke reduces visibility in real events, movement can slow even further as people hesitate, misjudge steps, or adjust their speed.

Human behaviour can lead to delays

Human behaviour is one of the biggest sources of delay in high-rise evacuations. People rarely act immediately when an alarm sounds. They pause, look for confirmation, check conditions, gather belongings, or coordinate with family members.

These early minutes are consistently some of the costliest when evacuating from tall buildings.

Studies of the World Trade Center evacuations show the more cues people saw – smoke, shaking, noise – the more they sought extra information before moving. That search for meaning adds delay. People talk to colleagues, look outside windows, phone family, or wait for an announcement. Ambiguous cues slow them even further.

In residential towers, families, neighbours and friend-groups naturally try to evacuate together. Groups tend to form wider steps, or group together in shapes that reduce overall flow. But our research shows when a group moves in a “snake” formation – one behind the other – they travel faster, occupy less space, and allow others to pass more easily.

These patterns matter in high-rise housing, where varied household types and mixed abilities make moving in groups the norm.

Why stairs aren’t enough

As high-rises grow taller and populations age, the old assumption that “everyone can take the stairs” simply no longer holds. A full building evacuation can take too long, and for many residents (older adults, people with mobility limitations, families evacuating together) long stair descents are sometimes impossible.

This is why many countries have turned to refuge floors: fire- and smoke-protected levels built into towers as safe staging points. These can reduce bottlenecks and prevent long queues. They give people somewhere safe to rest, transfer across to a clearer stair, or wait for firefighters. Essentially, they make vertical movement more manageable in buildings where continuous descent isn’t realistic.

Alongside them are evacuation elevators. These are lifts engineered to operate during a fire with pressurised shafts, protected lobbies and backup power. The most efficient evacuations use a mix of stairs and elevators, with ratios adjusted to the building height, density and demographics.

The lesson is clear: high-rise evacuation cannot rely on one tool. Stairs, refuge floors and protected elevators should all be made part of ensuring vertical living is safer.

Erica Kuligowski is affiliated with the Society of Fire Protection Engineers (SFPE) as a Section Editor for their Handbook of Fire Protection Engineering (Human Behaviour Section) and as a member of the Board of Governors for the SFPE Foundation. From 2002 to 2020, Erica worked as a research engineer and social scientist in the Engineering Laboratory of NIST, where she contributed to NIST’s Technical Investigation of the 2001 WTC Disaster and received US government funding to study occupant evacuation elevators.

Ruggiero Lovreglio receives funding from the Ministry of Business, Innovation and Employment (New Zealand), Royal Society Te Apārangi (New Zealand) and NIST (USA)

Milad Haghani does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. The Hong Kong high-rise fire shows how difficult it is to evacuate in an emergency – https://theconversation.com/the-hong-kong-high-rise-fire-shows-how-difficult-it-is-to-evacuate-in-an-emergency-270774

Source: The Conversation – Global Perspectives – By Ehsan Noroozinejad, Senior Researcher and Sustainable Future Lead, Urban Transformations Research Centre, Western Sydney University

At least 44 people have died and more than 270 are missing after a major fire engulfed an apartment complex in Hong Kong’s Tai Po district. The fire, which swept through multiple high-rise towers, is still burning.

The exact cause of the blaze, which broke out just before 3pm local time on Wednesday, is still unknown. Hong Kong Police have arrested three construction company executives on suspicion of manslaughter.

The apartment blocks are 31 stories tall. Opened in 1983, they were undergoing renovations at the time of the blaze, and were covered in bamboo scaffolding and green protective mesh.

Bamboo scaffolding has been a feature of the city for centuries. But why? The answer is part history, part engineering and part economics.

But the recent tragedy has sharpened the focus on fire safety, and when and where bamboo should be used.

A fast-growing grass

Bamboo is a fast-growing grass with hollow, tube-like stems (known as “culms”). Those tubes give it a high strength-to-weight ratio. A pole is light enough to carry up a stairwell, yet strong enough, when braced and tied correctly, to support platforms and workers.

Crews lash poles together in tight grids and tie them back to the buildings with brackets and anchors. Properly designed, a bamboo scaffold can resist wind and working loads.

Hong Kong’s Buildings Department and Labour Department publishes clear guidelines on the design and construction of bamboo scaffolds.

Bamboo scaffolding is also used in parts of mainland China, India, and across Southeast Asia and South America.

A cheap and flexible material

There are three main reasons why bamboo scaffolds are used in Hong Kong.

First, speed. An experienced team can “wrap” a building quickly because poles are light and can be cut to fit irregular shapes. That matters in tight streets with limited crane access.

Second, cost. Bamboo is a fraction of the price of metal systems, so contractors can keep bids low. The material is also easy to source locally, which keeps routine repairs and repainting within budget.

Third, tradition and skills. Bamboo scaffolding features in a famous piece of Chinese art, Along the River During the Qingming Festival, painted by Zhang Zeduan who lived between 1085 and 1145. Hong Kong still trains and certifies bamboo scaffolders, and the craft remains part of the city’s construction culture.

These factors explain why bamboo has remained visible on the city’s skyline even as metal systems dominate elsewhere.

Unlike metal made in blast furnaces, bamboo also grows back, and turning a stalk into a pole takes little processing. This means its overall climate impact is smaller.

What are the risks?

There are two main risks of bamboo scaffolding.

The first, as this tragedy in Hong Kong highlights, is fire.

Dry bamboo is combustible, and the green plastic mesh often draped over scaffolds can also quickly burn.

In the Tai Po fire, footage and reports indicate the fire quickly raced up the scaffolding and mesh, and across the facade of the buildings.

This is why there are calls for non-combustible temporary works on occupied towers – or at minimum, flame-retardant nets, treated bamboo, and breaks in the scaffold so fire can’t easily jump from bay to bay.

The second risk of using bamboo scaffolding is related to variability and weather.

Bamboo is a natural material, so strength varies with species, age and moisture. Lashings can loosen and storms are a common risk.

Hong Kong’s updated guidelines and code try to manage this with material rules (such as age, diameter and drying), mandatory ties to the structure, steel brackets and anchor testing, and frequent inspections – especially before bad weather.

A shift to metal

In March 2025, Hong Kong’s Development Bureau directed that metal scaffolds be adopted in at least 50% of new government public-works building contracts. It also encouraged metal use in maintenance where feasible.

Subsequent government replies to the Legislative Council in June and July reiterated the 50% requirement and described a progressive transition based on project feasibility.

Private projects may still use bamboo under existing codes. But for public works the baseline is now metal, signalling a move toward non-combustible systems.

The lesson from Hong Kong is not that bamboo is “good” or “bad” for scaffolding – it’s about context. It has clear advantages for small-scale, short-duration, ground-anchored work where streets are tight and budgets are lean. But on tall, occupied residential blocks, especially with mesh-wrapped facades, its fire risk and variability demand much stronger controls.

Bamboo scaffolding helped build Hong Kong’s skyline because it was fast, clever and affordable. The science behind fire and the realities of high-rise living now demand a tighter line: use the right tool for the job, and when the risks climb, switch to non-combustible systems.

That way the city can honour a proud craft, while keeping people safe in the homes those scaffolds surround.

Dr Ehsan Noroozinejad has received funding from both national and international organisations. His most recent funding on integrated housing and climate policy comes from the Australian Public Policy Institute (APPI). He also serves on the Executive Committee of the Early- and Mid-Career Academic and Practitioner (EMCAP) Network at Natural Hazards Research Australia, the Australian government-funded national centre for natural hazard resilience and disaster risk reduction.

– ref. Why is bamboo used for scaffolding in Hong Kong? A construction expert explains – https://theconversation.com/why-is-bamboo-used-for-scaffolding-in-hong-kong-a-construction-expert-explains-270780

Source: The Conversation – Global Perspectives – By Nathan Garland, Lecturer in Applied Mathematics and Physics, Griffith University

OpenAI chief executive Sam Altman – perhaps the most prominent face of the artificial intelligence (AI) boom that accelerated with the launch of ChatGPT in 2022 – loves scaling laws.

These widely admired rules of thumb linking the size of an AI model with its capabilities inform much of the headlong rush among the AI industry to buy up powerful computer chips, build unimaginably large data centres, and re-open shuttered nuclear plants.

As Altman argued in a blog post earlier this year, the thinking is that the “intelligence” of an AI model “roughly equals the log of the resources used to train and run it” – meaning you can steadily produce better performance by exponentially increasing the scale of data and computing power involved.

First observed in 2020 and further refined in 2022, the scaling laws for large language models (LLMs) come from drawing lines on charts of experimental data. For engineers, they give a simple formula that tells you how big to build the next model and what performance increase to expect.

Will the scaling laws keep on scaling as AI models get bigger and bigger? AI companies are betting hundreds of billions of dollars that they will – but history suggests it is not always so simple.

Scaling laws aren’t just for AI

Scaling laws can be wonderful. Modern aerodynamics is built on them, for example.

Using an elegant piece of mathematics called the Buckingham π theorem, engineers discovered how to compare small models in wind tunnels or test basins with full-scale planes and ships by making sure some key numbers matched up.

Those scaling ideas inform the design of almost everything that flies or floats, as well as industrial fans and pumps.

Another famous scaling idea underpinned the boom decades of the silicon chip revolution. Moore’s law – the idea that the number of the tiny switches called transistors on a microchip would double every two years or so – helped designers create the small, powerful computing technology we have today.

But there’s a catch: not all “scaling laws” are laws of nature. Some are purely mathematical and can hold indefinitely. Others are just lines fitted to data that work beautifully until you stray too far from the circumstances where they were measured or designed.

When scaling laws break down

History is littered with painful reminders of scaling laws that broke. A classic example is the collapse of the Tacoma Narrows Bridge in 1940.

The bridge was designed by scaling up what had worked for smaller bridges to something longer and slimmer. Engineers assumed the same scaling arguments would hold: if a certain ratio of stiffness to bridge length worked before, it should work again.

Instead, moderate winds set off an unexpected instability called aeroelastic flutter. The bridge deck tore itself apart, collapsing just four months after opening.

Likewise, even the “laws” of microchip manufacturing had an expiry date. For decades, Moore’s law (transistor counts doubling every couple of years) and Dennard scaling (a larger number of smaller transistors running faster while using the same amount of power) were astonishingly reliable guides for chip design and industry roadmaps.

As transistors became small enough to be measured in nanometres, however, those neat scaling rules began to collide with hard physical limits.

When transistor gates shrank to just a few atoms thick, they started leaking current and behaving unpredictably. The operating voltages could also no longer be reduced with being lost in background noise.

Eventually, shrinking was no longer the way forward. Chips have still grown more powerful, but now through new designs rather than just scaling down.

Laws of nature or rules of thumb?

The language-model scaling curves that Altman celebrates are real, and so far they’ve been extraordinarily useful.

They told researchers that models would keep getting better if you fed them enough data and computing power. They also showed earlier systems were not fundamentally limited – they just hadn’t had enough resources thrown at them.

But these are undoubtedly curves that have been fit to data. They are less like the derived mathematical scaling laws used in aerodynamics and more like the useful rules of thumb used in microchip design – and that means they likely won’t work forever.

The language model scaling rules don’t necessarily encode real-world problems such as limits to the availability of high-quality data for training, or the difficulty of getting AI to deal with novel tasks – let alone safety constraints or the economic difficulties of building data centres and power grids. There is no law of nature or theorem guaranteeing that “intelligence scales” forever.

Investing in the curves

So far, the scaling curves for AI look pretty smooth – but the financial curves are a different story.

Deutsche Bank recently warned of an AI “funding gap” based on Bain Capital estimates of a US$800 billion mismatch between projected AI revenues and the investment in chips, data centres and power that would be needed to keep current growth going.

JP Morgan, for their part, has estimated that the broader AI sector might need around US$650 billion in annual revenue just to earn a modest 10% return on the planned build-out of AI infrastructure.

We’re still finding out which kind of law governs frontier LLMs. The realities may keep playing along with the current scaling rules; or new bottlenecks – data, energy, users’ willingness to pay – may bend the curve.

Altman’s bet is that the LLM scaling laws will continue. If that’s so, it may be worth building enormous amounts of computing power because the gains are predictable. On the other hand, the banks’ growing unease is a reminder that some scaling stories can turn out to be Tacoma Narrows: beautiful curves in one context, hiding a nasty surprise in the next.

Nathan Garland does not work for, consult, own shares in or receive funding from any company or organisation that would benefit from this article, and has disclosed no relevant affiliations beyond their academic appointment.

– ref. Can bigger-is-better ‘scaling laws’ keep AI improving forever? History says we can’t be too sure – https://theconversation.com/can-bigger-is-better-scaling-laws-keep-ai-improving-forever-history-says-we-cant-be-too-sure-270448

Source: The Conversation – Global Perspectives – By Matthew Sussex, Associate Professor (Adj), Griffith Asia Institute; and Fellow, Strategic and Defence Studies Centre, Australian National University

A flurry of recent diplomatic activity has seen two competing peace plans for Ukraine emerge.

The first, widely touted as a US plan, was apparently hashed out between Kremlin insider Kirill Dmitriev and Steve Witkoff, President Donald Trump’s Russia point-man.

The second, hurriedly drafted by the United Kingdom, France and Germany, is based on the 28 points in the US plan, but with key modifications and deletions.

Following the release of the US plan, Trump accused Ukraine of showing “zero gratitude” for US assistance in the war effort, and demanded Kyiv accept the terms by Thanksgiving in the United States – November 27 – or face being cut off from US intelligence sharing and military aid.

Unlike the US plan, the European counter-proposal places the blame for the war squarely at Russia’s feet. It proposes freezing Russian assets until reparations are made by Moscow. It also seeks to freeze the conflict in place, leaving the question of which party retains which part of Ukraine contingent on subsequent negotiations.

Speaking about the peace proposals, European Commission President Ursula von der Leyen made it clear the European Union was committed to several key positions:

• that Ukraine’s borders cannot be altered by force
• there cannot be limitations on Ukraine’s armed forces that would leave it vulnerable, and
• the EU needed to have a seat at the table in any agreement.

Comparing the two plans, it is clear Russia and Europe remain as far apart as ever on Ukraine’s future. That much is unsurprising.

What should be more shocking to Western observers is just how much the US plan echoed Russian demands that have remained largely unaltered since President Vladimir Putin’s full-scale invasion of Ukraine in early 2022.

Lacking logic and specifics

Put simply, the US plan would have had as much credibility if it had been written in crayon.

For starters, it has wording that appears to make more sense in Russian than English (or perhaps AI-translated English).

And it seems more focused on bringing about a new era of friendly Russia-US economic cooperation than a serious attempt to resolve Europe’s biggest land war since the Second World War.

Typical of Trumpian robber-baron foreign policy, the document foresaw large cash grabs for the US, amounting to little more than attempts at extortion.

In return, Ukraine was offered a murky NATO-style security guarantee that could be reneged upon under flimsy pretexts.

The plan also demanded:

• large territorial concessions from Kyiv
• a limited army
• a pledge enshrined in Ukraine’s constitution that it would never to join NATO, and
• a promise to hold elections in 100 days.

And while it expected Ukraine to strategically emasculate itself, the document made only vague suggestions about what Russia is “expected” to do, with no means of enforcement.

No multinational force was put forward to monitor the peace. And Ukraine was required to give up key defensive positions by ceding the territory it still controls in the Donbas region to Russia. That would leave the centre of the country defenceless against future Russian attacks.

Accepting those terms, as originally written, would be politically suicidal for Ukrainian President Volodomyr Zelensky. This was obvious in his sombre message that the plan forced Ukraine to choose between its dignity and continued US support.

A point-by-point breakdown

A closer look at just a selection of its key points illustrates just how bizarre the plan is.

• Point 4 calls for “dialogue” between Russia and NATO, mediated by the US. That’s odd, since the US is a member of NATO.

• Point 7 requires NATO to include a provision in its statutes that Ukraine will not be admitted. But the main purpose of NATO is that membership is open to all.

• Point 9 says “European fighter jets” will be stationed in Poland, but doesn’t mention the American F-35s currently there.

• Point 10 states that if Ukraine launches a missile “without cause” at St Petersburg or Moscow (strangely implying it’s fine to hit Smolensk or Voronezh, for instance) – then Kyiv loses its US security guarantee.

• Point 13 says Russia will be invited to rejoin the G8 (the group now known as the G7 after Russia was expelled in 2014). But it says nothing about whether the other six members would agree to that.

• Point 16 requires Russia to enshrine in law a policy of non-aggression towards Ukraine. However, it had already done so several times in the past, yet still invaded Ukraine in 2022.

• Point 22 foresees a demilitarised zone in parts of Donetsk that Russian troops will not be able to enter. How to enforce that is left unspecified.

• Point 26 gives everyone involved in the conflict full amnesty for their actions, including numerous alleged war criminals.

• Point 27 establishes a “Peace Council” that would be overseen by Trump, similar to the “Board of Peace” envisioned in the Gaza peace plan, also headed by Trump. This gives him the ability to determine whether the agreement is being violated (and, crucially, by whom).

Where to next?

Ukrainians have been sold a diplomatic lemon before. In 1994, Ukraine signed the Budapest Memorandum, in which Kyiv agreed to give up the nuclear weapons it still held from the Soviet era, in return for commitments by Russia and the US that its sovereignty and borders would be respected.

Just as the current US plan has been rebuffed by Kyiv, there is no hope of the European alternative being endorsed by the Putin regime. Indeed, it has already been rejected by one of Putin’s senior advisers.

Where does this leave the peace process? US Secretary of State Marco Rubio has already walked back the US plan from a concrete set of demands to a “living, breathing document”, and hinted at great progress in negotiations with Ukraine.

European and Ukrainian stakeholders have also made approving noises, knowing that if the White House loses interest, securing peace will be much harder.

Yet it’s below the surface that the real soul-searching will be happening, in Ukraine, as well as the broader West. Once again, the Trump administration has proven it is more interested in long-term deals with autocrats than achieving just and lasting resolutions to security crises.

That alone should give US allies pause, and not just in Europe. For those nations, it’s one thing to doubt Putin’s motives. But it’s another thing entirely to now have to doubt America’s as well.

Matthew Sussex has received funding from the Australian Research Council, the Atlantic Council, the Fulbright Foundation, the Carnegie Foundation, the Lowy Institute and various Australian government departments and agencies.

– ref. How the Trump administration tried to sell Ukraine a diplomatic debacle – https://theconversation.com/how-the-trump-administration-tried-to-sell-ukraine-a-diplomatic-debacle-270561