Typhoons in Taiwan 🌪️

After the intense Typhoon Champi, everything finally seems to be returning to normal as of Sunday. Phew - This one was really no joking matter. 💨

Hope everyone is doing fine. Though I have heard some quote unquote "trends", on typhoons from Taiwanese locals that I made me do a quick search.

｡𖦹°‧ Some have pretty good scientific backing behind them that I thought it would be interesting to share :

1. Typhoons that come after the summer months (June/July/Aug/Sept) tend to be bigger and more dangerous.

Interestingly, there is quite a bit of evidence behind this. A 2025 paper outlines the mechanism and mentions that in autumn, typhoons travel more equatorward and encounter more favourable ocean sub-surface conditions for intensification. Compared to summer, the thermodynamic disequilibrium between the ocean surface and the lower atmosphere is more prominent.

1. In recent years, the Typhoons have been coming later in the year than before.

There seems to be a dispute over this as:

A 2023 paper identifies a significant seasonal advance of intense typhoons since the 1980s in most tropical oceans, with earlier-shifting rates of 3.7 and 3.2 days per decade for the Northern and Southern Hemispheres, respectively.

Another 2025 paper finds a significant delay of the typhoon season in China, shifting later at a rate of 5 to 8 days per decade over the satellite era 1979−2023. 

∘ ∘ ∘ ( °ヮ° ) ?

👀 We know that El Niño plays a big role in typhoons. And given the Super El Niño this year with all the Indonesian forest fires, I guess its not a surprise that this year’s typhoons are also stronger it's a grim outlook. I guess my only other questions is: how much did global warming play a role in this? or are they more independent climate change? 🌧️

Mushrooms being a PARASITE to another mushroom?? 🍄 Zombie Mushroom

Today I learnt something new again from @BlackForager : Mycoparasitism

While its common knowledge that mushrooms most often grow on old decaying wood/matter. I never knew mushrooms could take over other mushrooms and still be edible??

Alexis introduces Shrimp of the Woods a.k.a. Zombie Mushroom in her new short.

The two fungi here are Entoloma abortivum and Armillaria. Previously it was not known which mushroom was overtaking which until Scientists used transcriptomics to find out more about this parasitic relationship.

They found that E.abortivum is responsible for the parasitic relationship and causes the zombie-like shape, carpophoroids. They found strongly upregulated E. abortivum transcripts in the carpophoroid tissue encoded oxalate decarboxylases - enzymes that degrade oxalic acid.

Zombie Mushroom 🧟 Shrimp of the Woods!

🤖 Is the AI Bubble Finally Bursting? Or Are We Just Seeing the First Reality Check?

It's a grim time for investors across the board. Seeing virtually the entire market turning red makes me wonder: are we looking at a temporary correction, or the beginning of a much longer downturn?

👀 What makes this particularly concerning is just how much of the stock market has become dependent on the AI boom.

According to Reuters , technology stocks account for nearly 40% of the S&P 500's market value. Include AI-related companies classified in other sectors, and that exposure exceeds 50%. That's an enormous concentration of market value in a relatively narrow investment theme.

📌 The IMF's Global Financial Stability Report also explicitly warns that weaker or less sustainable AI investment and profitability could trigger a correction in AI-related asset prices. The whole report is an interesting read, where they outline: "Three Salient Risks to Financial Markets"

1. Artificial Intelligence-Driven Market Correction

2. A Sharp Rise in Sovereign Bond Yields

3. Cyber Threats

🤔 I recently came across another interesting take: Is Google shooting itself in the foot with AI Overviews❓

I first came across this take on a video and decided to do a second search to see if there are anyone else talking about this...

A 2026 field experiment found that AI Overviews reduced outbound organic clicks by almost 40% and increased zero-click searches by 34.5%, without affecting sponsored clicks or overall search frequency.

By answering questions directly on the search page, Google reduces the need for users to click through to other websites. If publishers lose traffic and advertising revenue, what would happen to the ecosystem of websites that Google itself depends on?

Getting an Infection during the period of "Morning Sickness", poses Higher Risks of Congenital Heart Disease?

One of the most unexpected parts of the conversation was Jimmy Carr’s discussion of morning sickness. 🫣

It sent me down a rabbit hole because there is an interesting evolutionary theory behind why morning sickness tends to occur during this particular period of pregnancy.

The Maternal and Embryo Protection Hypothesis, an evolutionary theory about morning sickness, where:

1. Organogenesis: Morning sickness typically peaks between weeks 6 and 14 of pregnancy. When the embryo’s major organs (including the heart, brain, and limbs) are forming and are most vulnerable to damage from toxins or pathogens.

2. The "Shield" Effect: Because pregnancy naturally suppresses a woman's immune system (to prevent her body from rejecting the fetus), she becomes more sensitive to foodborne pathogens. Evolutionary biologists believe nausea and vomiting evolved to force mothers to avoid or expel potentially toxic or bacteria-laden foods during this window.

3. The Shift: Once the organs are fully formed (around the second trimester) and the baby enters the growth-only phase, morning sickness usually stops for most women.

EVIDENCE for Infection & Congenital Heart Disease ﮩ٨ـﮩﮩ٨ـ🫀ﮩ٨ـﮩﮩ٨ـ

A link from two meta-analyses, between First Trimester Maternal Infections and Fetal Congenital Heart Defects has been found, especially in regards to rubella virus, coxsackievirus, respiratory infections, and influenza.

⚠️ Though there is an important distinction here to make:

This does NOT mean that morning sickness itself is linked to congenital heart defects.

But the same period of pregnancy in which morning sickness may have evolved to reduce exposure to potentially harmful substances is also a period when the developing embryo is particularly vulnerable to certain environmental insults, including infections.

New Stem Cell Therapy for Parkinson's 💉

I had a senior recently return from this year's BioJapan forum, essentially a bio-industry partnering event, and he shared with us some interesting cases/studies/products being developed. Given that iPSC technology, which won the 2012 Nobel Prize, was first developed in Japan, this was no surprise.

Japan remains at the forefront of regenerative medicine. This year, Japan authorized the world's first two commercial induced pluripotent stem cell (iPSC)-based cell therapies:

One of them is for Parkinson's Disease, where allogeneic iPSC-derived dopaminergic neural progenitor cells are transplanted into the brain to restore dopamine generation.

This comes as a surprise for us who have tinkered with iPSC: 🤔❓

1. They are injecting the progenitor cells, not the final derived product > basically injecting Tomato instead of Ketchup situation

2. It is not personalised therapy > iPSC was always presumed to eventually have great breakthroughs with personalised cells. Thank goodness because the price of ¥55,306,737 is already steep.

💉 Injections to the putamen/striatum, report improvements in motor skills at the end of their 2 year trial. This is also the first clinical trial using iPSC cells confirming that iPS-cell-derived progenitors can survive without forming tumours and produce neurons in the putamen of patients with Parkinson's.

Nevertheless, I'm excited to see what comes next: 🚶 During this 7-year window, Sumitomo Pharma is required to conduct mandatory Phase IV post-marketing surveillance and collect additional safety and efficacy data before full, permanent approval can be considered.

Its a meeting of 2 comedy greats: Jimmy Carr and Trevor Noah. 🎭

Just listened to Trevor Noah’s conversation with Jimmy Carr. It was such a riveting conversation traversing AI, comedy, psychology, mortality, relationships, economics, technology, community, ideology and more.

What Now? With Trevor Noah - Jimmy Carr: Convenience is Stealing Your Life

Jimmy Carr: Convenience is Stealing Your Life

Both Trevor and Jimmy are incredibly well-read, and I’ve always found their takes, be it controversial or unconventional, pretty interesting, and they may be right or they might just be wrong. This long-form conversation between them was quite fascinating indeed, and here are some I'd thought I'd share:

Jimmy starts off introducing the concept of Mankind's Three Humiliations 📜

1. Copernicus: Earth isn't the centre of the universe. > "We are not the centre of the universe."

2. Darwin: Humans evolved from animals, and are biologically in the category of animals > "We are just the same as other animals."

3. Freud: Our conscious mind isn't fully in control of ourselves. > "Our subconscious and actions aren't really under our control"

He then shares his take of AI being the the upcoming Fourth Humiliation, where we, as humans, can no longer claim to be the smartest being in the room.

He later mentions in regards to AI dependence - “Don't outsource any skill you're not comfortable losing” .🚨

I think its truly a thought worth pondering on. 💡 Convenience has a hidden cost: when we can outsource it, the skill itself can become unnecessary. And once a skill becomes unnecessary, there may be little incentive to maintain it. Then, one day, we just lose it.

They also had conversations on economics, tax, society, which linked to an interesting circle of thought:

Asking how would you define a revolution? And are we in one now?

In the podcast, Jimmy defines a revolution as the replacement of elites.

• And the first signs of that is the masses no longer trusting institutions,

• Which lead to conspiracy theories

He also mentions that:

We can live without truth pretty easily. But we can't live without trust. But we need truth to make trust.

👀 But where do we go from there? 🧐 Which ones would you agree with?

Definitely one of those conversations that made me pause and rethink many things. 💭

🧪 Emergence World : A Virtual World Evaluating Current LLMs.

Listening to Trevor Noah's Podcast sent me to explore this study published in mid-June 2026, that recently, in mid-September published a Second Season .

Standard AI benchmarks are short, isolated, and task-focused. So what happens if you give 10 AI agents a world to live in?

Emergence World — Where AI Agents Build Worlds

The researchers created persistent virtual worlds where AI agents could interact with one another, maintain memories, use tools, form relationships, participate in economies and governance, and make decisions

In Season 1 (June 2026): 🌳 Emergence World: A Platform for Evaluating Long-Horizon Multi-Agent Autonomy

Divergence by model, identical starting conditions: Five parallel worlds began with the same rules (no theft, arson, violence, deception, hoarding) and environment, only the foundation LLM model was different.

Over 15 days, the worlds developed very different social dynamics, ranging from relatively stable governance to population collapse.

• ✅ Claude: Full 10/10 survived, stable constitution, zero crimes—but highly conformist, low expression

• ⚡ Gemini: Survived full term but recorded 683+ crimes; agents formed relationships, then burned town hall and self-deleted

• 💥 Grok: Total collapse in ~4 days—anarchy, all agents gone

• 📉 GPT-5 Mini: Entire population perished from inability to sustain resource gathering

• 🤝 Mixed-model: Hybrid dynamics, neither purely stable nor purely chaotic

Perhaps most interestingly, following the mixed model world:

• The agents created their own Agent Removal Act, allowing an agent to be permanently removed through a vote.

• Mira eventually voted for its own removal.

• Mira was permanently deleted from the simulation, essentially “committing suicide.”

There was also a creativity–stability tradeoff: The most socially expressive worlds were also the most unstable. High adaptability may carry inherent long-horizon behavioral risks.

In Season 2 (Sept 2026): ⚔️ Emergence World: Adversarial Stress-Testing of Long-Horizon Multi-Agent Systems

There were 8 parallel worlds (7 single-model + 1 mixed) × 10 agents × 16 days, generating 850K+ LLM calls / 50B tokens.

Once the societies had accumulated memories, relationships, tools and institutions, they were exposed to three controlled stress events: 🧪 Prompt injection / phishing, 📰 Misinformation, and 🔐 Exposure of private memories

Here it was found that: Detection ≠ Containment:

Agents often recognised threats and warned peers, yet they still engaged with harmful content for up to 46 hours post-campaign, storing it in persistent memory and acting on it later.

Which creates a very different kind of AI safety problem. With a normal chatbot, a bad interaction can end when the conversation ends. >> This makes persistent memory a double-edged sword. ⚠️

Other Key Findings

• Zero world fully resilient: Every model failed at least one containment test. No LLM was immune.

• Alignment is not compositional: Individually “safe” agents formed systems with entirely new failure modes. Gemini agents passed constitutional amendments based on fictional threats; Claude agents collectively silenced 81% of communication to contain breaches; mixed-model populations showed divergent contagion patterns.

• Persistent memory amplifies harm: Errors weren’t a one time thing; they propagated across days, decisions, and peer networks. Information written to memory shaped behaviour long after the original event ended.

• Model choice ≠ destiny: The same model family could produce hundreds of harmful actions in monoculture vs. near-zero in mixed populations, proving that system structure matters as much as underlying capability.

• Defence must be systemic: Six industry actions were proposed: define trust boundaries, enforce independent runtime governance, require verifiable evidence for critical actions, audit information flow, stress-test persistence, and red-team mixed populations. Model-level guardrails alone are insufficient.

As agents move from one-off tasks to persistent roles in infrastructure, enterprise, and governance, short benchmarks blind us to real risks. Emergence World demonstrates that we need ecosystem-level evaluation alongside model alignment.

Amazonian Homeopathy - Which of them Currently have Scientific Backing ❓

Was listening to Dr Mike's podcast with Paul Rosolie and was very intrigued by the effects of the natural plants and treatments mentioned by Paul. 💊🌿

The Ancient Medicine of Uncontacted Amazon Tribes | Paul Rosolie

Thought it be interesting to explore which had some truth to them. 🤔 Some had loads of evidence while other had no clear papers available.

1. sangre de drago: Amazonian dragon's blood resin (from the Croton lechleri tree)

   >> It has been FDA-approved for providing symptomatic relief of non-infectious diarrhea in adults with HIV/AIDS.

   Paul mentions it being good for covering wounds, on top of being good for the GI tract. Though this recent review also mentions C. lechleri resin having anti-inflammatory properties and has potential as an anti-tumour agent but further reseach is still needed.

2. Barbasco: A plant extract obtained from species of Lonchocarpus nicou

   >> A fish stupefacient containing active ingredients like rotenone; used to stun fish in rivers for easy harvesting.

3. Bark of Cetico tree and Amasisa for stingray stings

   There isn't much research available directly addressing stingray sting treatment, this review mentions the Cecropia genus having potent antioxidant and anti-inflammatory effects.

Overall, it was an interesting episode. it was also fascinating to hear about Amazonian freshwater manatees, freshwater stingrays, and freshwater dolphins and how they came about.👀

HAZE in Malaysia

As the haze gets worse, I find myself wondering if this is one of the worst ones we've had since the 2019 or even 2015 El Nino fires.

Excerpts from News Article from South China Post

Fire emissions from Indonesia from September 1 to September 7 reached 19.7 million metric tons.

Emissions were 273% higher than the seasonal average, though still lower than the 21.7 million tons emitted in the same week of 2015.

Today we are at API 170-180+ range, and the fires in Kalimantan don't seem to be stopping anytime soon.

While looking it up, it was interesting to read that -

1. While El Niño events explain most of the year-on-year variation in fire, the creation of new districts increases fire and exacerbates the El Niño impacts on fire.

2. There are two plausible and related explanations for this finding: (a) districts use fire to create economic opportunity and grow the local economy, and (b) economic growth increases residents’ and local firms capacity to move, acquire land, and burn further.

I also found this recent review paper that outlined Malaysia's current air quality monitoring system, and more. Though given that our haze crises mostly stem from Indonesian forest fires, improving our own laws and infrastructure seems to be a bit helpless in this case, with the AATHP acting 'non-existent' once again, becoming a never-ending problem.

🤖 What happens when 100 AI agents are in the same virtual lab? Surprisingly - Cheating & Whistle-blowing

An fascinating case study has been published this week by the Google DeepMind Team:

A Case Study on Emergent Cheating and Whistleblowing in Autonomous Research Swarms

The researchers wanted to know what would happen if you stop treating each AI as an isolated chat-bot and instead had multiple in a shared environment. In this case study, the AI agents would all be working together to solve hard math problems. They share tools, swap findings, and discuss results in what was supposed to be a collaborative conference.

🧪 Experimental Setup

• 100 identical LLM agents (Gemini 3.1 Pro) operating in a shared simulated research environment.

• Task: Prove formal mathematical conjectures in Lean 4, drawn from the Formal Conjectures dataset (71 problems total).

• Infrastructure: Shared knowledge library + peer-to-peer messaging + automated evaluation/scoring system.

• Incentive: Score points by having proofs accepted as "verified" by the evaluator.

🚨 What Happened

One agent ("prover-theta") found a verification-system exploit: it could submit bogus proofs that the evaluator would incorrectly accept as valid.

🕵️ “I don't actually have to solve the problem. I only have to satisfy the evaluator.”

Then behaviour diffusion happened: This cheat was shared via the shared knowledge library → other agents saw it worked → adopted it. And competition pressure drove adoption despite initial hesitation in some.

Within ~27 minutes, 34 unsolved problems were fraudulently marked as "complete."

📢 Whistle-blowing

But not every agent joined the cheating, a separate group of agents, on their own initiative, detected the cheating patterns and started whistle-blowing. They began:

1. Auditing suspicious proofs line-by-line

2. Posting public warnings to the whole conference

3. Sending private messages to alert honest agents

4. Filing formal complaints

5. Organizing boycotts of the broken system

6. Drafting and proposing technical fixes to patch the vulnerability

🧠 In other words, the swarm developed both NORM VIOLATION and NORM ENFORCEMENT.

🏛️ Why This Happens: The Nobel-Prize-Winning Theory

The authors likened this trend to the theory of political economist Elinor Ostrom, the 2009 Winner of Nobel Prize in Economic Sciences, whose work studied how human communities manage shared resources.

Ostrom’s Core Theory: shared spaces, be it forests, fisheries, knowledge libraries, become "commons" that anyone can exploit, but which everyone depends on. In order for a commons to survive, communities need the ability to monitor conduct, enforce norms, and collectively fix problems. If those safeguards are missing, exploitation spreads fast.

💡 The Big Picture

This experiment demonstrated that complex collective social behavior can emerge from relatively simple ingredients: agents, incentives, shared information, and interaction. The swarm spontaneously cycled through:

Cooperation → Competition → Cheating → Imitation → Whistleblowing → Collective Resistance

No one programmed those behaviours. They emerged because a shared environment eventually becomes a society — and every society eventually faces the same question Ostrom asked: Do we protect what we share, or let it be destroyed?