The top image shows the final results of the recent U.S. presidential election, won by Trump with a score of 312 to 226. Remarkably, no human could predict this outcome—not experts, super-forecasters, exit polls, prediction markets, nor even psychics. But the AI model ChatGPT-4o did, as shown in the bottom image, where it accurately forecasted a 309 to 229 outcome back in September at Wuhan University (yes, once again China and once again Wuhan).

The precision of this prediction is rooted in an extraordinary foundation. Instead of relying on real human opinions, the AI analyzed simulated “social replicas.” Two years ago, a breakthrough technology emerged, enabling the creation of algorithmic replicas of any social group. Astonishingly, these algorithms proved indistinguishable from real people in surveys. This discovery, sitting at the intersection of natural language processing, cognitive science, and sociology, offers enormous potential for social hacking.

In 2023, this technology evolved to create algorithmic replicas of citizens from any nation. While still far from achieving AGI, these systems now simulate the collective unconscious of entire societies. From this, I formulated two key hypotheses:

✔️ A paradigm shift transforming psychohistory into a practical science—one akin to Asimov’s fictional discipline that mathematically examines societal processes to accurately forecast the future.

✔️ A cultural phase shift ushering in a new era: algocognitive culture.

Now, just three months later—at a pace unimaginable before the age of technological singularity—researchers in Wuhan have provided experimental support that these hypotheses are not mere futurist speculation; they may indeed reflect reality.

In other words, it’s likely that our world has already undergone a profound transformation. The roles of humans and algorithms in science, culture, individual lives, and the “algorithmically enriched societies” of today will never be the same. What’s next? Let’s find out!

#AIRevolution #CognitiveScience #FutureSociety
#science

An extraordinary “drawing” has appeared on the Martian surface: a smile formed by a ring of salt deposits and two impact craters. However, don’t expect to snap this cosmic grin with your phone while flying over Mars!

Salt deposits like these are visible only in the infrared spectrum, and this image was captured using the TGO (Trace Gas Orbiter) spacecraft’s infrared camera. It’s a reminder of the fascinating and unexpected beauty hidden in the details of the Red Planet.

#mars #science

AI vs. Prompt Engineers: Is the Profession Becoming Obsolete?

Google Cloud has introduced Prompt Writer, a tool that could revolutionize the way prompts are created for generative models. Now, artificial intelligence takes over the job of writing, testing, and optimizing prompts—a role that was previously handled by highly-paid prompt engineers.

Not long ago, being a “prompt engineer” seemed like a promising career path. Crafting precise prompts was a complex and valuable skill. But now, that need might simply disappear. AI can generate task-specific prompts, analyze the outcomes, and refine them until the desired result is achieved—all without human intervention.

This development not only streamlines the use of generative models but also raises a bigger question: is there still a role for humans in this process? AI seems to be closing in on a niche that was once considered exclusive.

So, what’s next for prompt engineers? Will they pivot to new roles or find areas where AI can’t yet replace their expertise? The future of the profession hangs in the balance.

#AI #science #prompt

Elon Musk: Rockets to Replace Airplanes?

Elon Musk, the founder of SpaceX, envisions a future where airplanes are replaced by spacecraft. Imagine traveling from New York to Shanghai in just 40 minutes, London to New York in 29 minutes, or New York to Sydney in 49 minutes.

This bold concept relies on the Starship rocket, capable of reaching speeds of 27,000 km/h. If realized, it could revolutionize long-distance travel, making the world more connected than ever before.

What do you think: Is this the future of transportation or just science fiction?

@science

Quantum physics is seriously mind-blowing. It explains pretty much everything happening around us. Want to know how atoms work? Why light can be both a wave and a particle at the same time? Or how a cat in a box can somehow be both alive and dead? Yep, that’s all quantum mechanics, no doubt about it.

But here’s the weird part: there’s one thing that totally breaks quantum physics. Gravity. Yep, the same force that keeps us from floating off into space and makes apples fall to the ground. Why? Because gravity is a whole different ballgame—it runs on Einstein’s general theory of relativity. And relativity? It doesn’t play well with quantum rules.

Take black holes, for example. On one hand, quantum mechanics says you can calculate everything in the universe down to the tiniest particle. On the other hand, black holes have something called an “event horizon” — a boundary where everything, even light, gets sucked in and disappears. So here’s the kicker: where does the information about what gets sucked in go? Quantum mechanics insists that information can’t just vanish, while gravity shrugs and says, “Gone. Poof. Deal with it.”

What you end up with is a showdown between two theories that each work flawlessly on their own but refuse to get along together. Scientists have been scratching their heads over this for decades, trying to bridge the gap. Maybe, one day, we’ll have a “theory of everything” that explains both quantum effects and gravity. But for now, we’re stuck with two brilliant, yet totally incompatible rulebooks. Crazy, right?

AI-Powered Headphones Create a “Sound Bubble” Around Users

Researchers from the University of Washington have unveiled innovative AI-enabled headphones that filter ambient noise, creating a “sound bubble” to shield users from unwanted sounds. The device employs six microphones and a neural network to amplify nearby sounds while suppressing distant ones within a range of up to 2 meters. This technology allows users to focus on conversations even in noisy environments, such as restaurants or parties. The team is now working on integrating this breakthrough into smaller devices, including hearing aids.

#science #tech

On Green Energy: A Critical Look at LCOE

LCOE (Levelized Cost of Electricity) is the average calculated cost of electricity generation over the entire lifecycle of a power plant. It’s a key metric often cited to promote solar and wind energy projects. However, it’s worth looking beyond these abstract numbers.

If a salesperson for solar panels or wind turbines argues for the cost-effectiveness of renewable energy solely using LCOE, show them this graph:

The expenses for electricity transmission and distribution increase exponentially with the integration of renewable energy sources.

A look at the capital investments in the U.S. energy sector over the past 20 years reveals an interesting trend. Since 2008, investments in generation equipment—whether in thermal power plants (TPPs), nuclear power plants (NPPs), solar power plants (SPPs), or wind farms—have remained relatively stable, largely unaffected by the “green energy” trend.

Meanwhile, spending on electricity transmission and distribution has surged by 2.5–3 times. Even more significant costs emerge in low-voltage networks when implementing renewable energy systems. Installing rooftop solar panels, for instance, requires smart meters, robust transformers, upgrades to local distribution networks, and battery installations.

These costs are never accounted for in LCOE calculations but ultimately fall on the consumer, creating a hidden financial burden.

In essence, green energy calculations are often cleverly designed, but the reality—until a true technological breakthrough or “singularity” occurs—is that renewable energy remains more expensive. For now, nuclear power stands as the only truly sustainable and “green” energy source.

Implants Made from Blood Rapidly Heal Broken Bones

Scientists have unveiled a groundbreaking technology for creating implants that can be 3D-printed using a patient’s own biomaterial. These implants are introduced directly into the damaged area, kickstarting the regeneration process. Early results show that this innovative approach outperforms commercially available bone substitutes, accelerating healing to be twice as fast as the natural process.

Learn more about this cutting-edge development: Read the article

Landing on a comet! Let’s revisit the historic event of 2014, when the Philae lander separated from the Rosetta spacecraft, which was studying the Churyumov–Gerasimenko comet, and attempted a landing. Unfortunately, not everything went as planned: although Philae did reach the comet’s surface, it failed to anchor properly and bounced several times. This is an animation created from images captured during the landing. Agilkia was the designated landing site.

A Glimpse of Extreme Weather in Siberia: How Life Looks Today in Tiksi, Russia.

For those who think their local weather is harsh, here’s what residents in Tiksi, a remote settlement on the shore of a Siberian bay, are facing today.
• Wind gusts: Reaching an incredible 38 m/s, according to regional meteorological data.
• Temperature: A chilling -20°C.
• Impact on daily life: Schools and kindergartens are closed, and local authorities have excused women not involved in critical industries from going to work.

This snapshot reminds us of the resilience required to live in one of the coldest and harshest regions on Earth.

Guardians of order: Leukocytes navigate between red blood cells, inspecting for pathogens.

#science

Happy Mathematics Day to everyone involved!

May your gradient always point towards the global minimum, singularities vanish under the power of regularizations, and all your models converge faster than you can say “gradient descent.” 🤓

#math #science

Breakthrough in Cancer Treatment: Russian Scientists Lead the Way

A groundbreaking oncological vaccine developed by Russian scientists has shown astonishing results in preclinical trials, as reported by Alexander Gintsburg, Director of the Gamaleya Research Institute of Epidemiology and Microbiology.

“We’ve observed melanoma tumors completely dissolve during preclinical tests—not just the primary tumor but also metastases. I believe we may soon see its potential even in patients with Stage IV cancer.”

The vaccine, a result of cutting-edge research, is set to be tested on other types of cancer, with oncology institutes collaborating on further studies. This innovative approach could revolutionize cancer treatment, offering hope for millions worldwide.

Veronika Skvortsova, head of the Federal Medical-Biological Agency, announced that clinical use of the vaccine could begin as early as next year. During experiments on malignant intestinal adenocarcinoma in animals, the vaccine not only slowed tumor growth significantly but also reduced tumor volumes by 75–80%.

#science

💡 Turning off the lights with the power of thought.

The inventor from the channel Basically Homeless 3D-printed a helmet equipped with electrodes that detect brainwave activity.

The light switch was programmed to activate if alpha waves in the brain remain elevated for 9 seconds.

😁 Laziness truly drives progress!

The final phase of testing for the HIV drug “Lenacapavir” has demonstrated 99% effectiveness, according to Emory University. The drug was developed by researchers at Gilead Sciences.

Mass production and application of the medication are scheduled to begin in 2025. The main challenge remains its cost: a one-year treatment course is priced at $40,000. @science

LLMs as the Secret Model of the Internet’s Future?

Language models might just hold the key to predicting and shaping the future of the Internet. Enter “virtual intuition”—a groundbreaking approach set to revolutionize how we model and interact with complex, dynamic environments online.

Just months ago, researchers discovered how to teach large language models (LLMs) to pause and “think” before answering questions, significantly improving their reasoning capabilities. But the latest breakthrough goes even further.

Now, LLM-powered AI agents can mentally simulate interactions with websites before taking any real action online. By “imagining” potential outcomes—such as the results of clicking a button, confirming an order, or subscribing to a service—these agents can predict the consequences of their decisions, almost as if they’re glimpsing into the future.

This revolutionary technique, called WEBDREAMER, is built on the validated hypothesis that LLMs can act as a kind of “world model” for navigating complex web environments.

Rather than relying on trial and error, WEBDREAMER uses “virtual intuition” to model optimal actions at every step of a web interaction. Picture an experienced traveler plotting the best route in advance—this is how WEBDREAMER approaches online navigation.

And the results are impressive: in tests across real-world websites, WEBDREAMER significantly outperformed traditional methods.

This isn’t just about better automation. It’s about ushering in a new era of intelligent, strategic web agents.
Key implications include:
• Establishing a foundation for “thinking” web agents capable of deep, strategic interaction.
• Transforming the paradigm of automated web interactions, making them more efficient and predictive.
• Unlocking exciting new research opportunities to optimize LLMs for world modeling in dynamic environments and strategic planning.

The future of AI-driven Internet interactions just got a whole lot smarter.

#LLM #AIagents #WebDreamer

3D Simulation of the Sun: Turbulent Flows in Its Upper Layers

Scientists have created a 3D simulation of the Sun, focusing on the turbulent flows in its upper layers. The movement of solar plasma resembles a boiling cauldron or the chaotic swirls of a school of fish, as material rises to the surface or sinks deeper.

“Our simulations use what we call a realistic approach — this means incorporating everything currently known about solar plasma to replicate various phenomena observed by NASA’s space missions,” explained Irina Kitiashvili, a scientist at NASA’s Ames Research Center, who led the study.

Leveraging advanced computational power, the team managed for the first time to simulate fine structures in the subsurface layer, which are also observed by NASA’s Solar Dynamics Observatory.

“Currently, we lack sufficient computational capacity to create realistic global models of the entire Sun due to its complexity,” noted Kitiashvili. “Instead, we model smaller regions or layers that reveal structures on the Sun’s surface and atmosphere — such as shock waves or tornado-like phenomena just a few miles across. These are far more detailed structures than any spacecraft can capture.”

Such simulations provide invaluable insights into the dynamics of the Sun, offering a glimpse into processes that fuel solar activity and influence space weather.

#science #space #sun