Loading video...

Video Failed to Load

Go Home

Google's new AI can predict flash floods 24 hours before they strike. How it works: > Uses Gemini to extract confirmed flood locations and times from global news > Builds a dataset of past events that never formally existed. > That dataset feeds a neural network > The neural...

rowancheung's profile picture

Rowan Cheung

593,963 subscribers

23,274 views • 4 months ago •via X (Twitter)

LIVE

Anya Rossi• Live Now

Private livecam show

0 Comments

No comments available

Comments from the original post will appear here

Related Videos

River Sentry Flood Warning is a Liberty Hill business that created flash flood warning systems. They warned people in Ingram this morning: "At Approximately 0332 AM, six of our RS-1A siren towers were involved in a flash flood intervention after detecting high water at Howdy's and River Run Campground near Ingram, Texas. All units activated, triggering the immediate evacuation of the camp ground and surrounding residents who were awoken by the towers. Three towers were damaged and/or lost and this is OK. We always place our equipment between sleeping quarters and the river to ensure water must encounter our sensors before it reaches people. This strategy was proven sound early this AM. We may have to replace some gear but so be it. It accomplished its purpose." FOX 7 Austin
0:57

River Sentry Flood Warning is a Liberty Hill business that created flash flood warning systems. They warned people in Ingram this morning: "At Approximately 0332 AM, six of our RS-1A siren towers were involved in a flash flood intervention after detecting high water at Howdy's and River Run Campground near Ingram, Texas. All units activated, triggering the immediate evacuation of the camp ground and surrounding residents who were awoken by the towers. Three towers were damaged and/or lost and this is OK. We always place our equipment between sleeping quarters and the river to ensure water must encounter our sensors before it reaches people. This strategy was proven sound early this AM. We may have to replace some gear but so be it. It accomplished its purpose." FOX 7 Austin

Chris Walker

23,137 views • 27 days ago

Climate change caused the catastrophic floods that tore through Central Texas over the last few days, killing at least 27 people, including nine children, and turning calm rivers into violent torrents, according to the media. At Camp Mystic in Kerr County, the Guadalupe River surged from about three feet to nearly 29 feet in just 90 minutes, sweeping away cabins, vehicles, and people with little or no warning. Climate change caused a warmer atmosphere, which holds more moisture, and unleashed it in increasingly intense bursts. Volumes of precipitation were extreme. They had less than a 0.1% chance of occurring in any given year, according to the New York Times. Texas climatologists warned that the frequency and severity of such events have already increased and could intensify by another 10 percent by 2036. In East Texas, “the number of days per year with at least two inches of rain or snow has increased by 20 percent since 1900,” noted the Times. But that tiny increase in precipitation doesn’t explain the floods or the deaths in Texas. Over the past century, global flood deaths declined by more than 80 percent. That happened not because nations reduced rainfall but because they have learned how to live with it. They built levees, dams, and drainage systems. They developed early warning systems and evacuated people before the water arrived. Kerr County in Texas failed to do any of these things. Despite its location in one of the most flood-prone regions in the United States, the county had no formal flood warning system in place. There were no sirens, no automated text alerts, no rapid evacuation protocol. The river rose, and families had no idea it was coming... Please subscribe now to watch the full video, read the full article, and support Public's award-winning reporting!
4:50

Climate change caused the catastrophic floods that tore through Central Texas over the last few days, killing at least 27 people, including nine children, and turning calm rivers into violent torrents, according to the media. At Camp Mystic in Kerr County, the Guadalupe River surged from about three feet to nearly 29 feet in just 90 minutes, sweeping away cabins, vehicles, and people with little or no warning. Climate change caused a warmer atmosphere, which holds more moisture, and unleashed it in increasingly intense bursts. Volumes of precipitation were extreme. They had less than a 0.1% chance of occurring in any given year, according to the New York Times. Texas climatologists warned that the frequency and severity of such events have already increased and could intensify by another 10 percent by 2036. In East Texas, “the number of days per year with at least two inches of rain or snow has increased by 20 percent since 1900,” noted the Times. But that tiny increase in precipitation doesn’t explain the floods or the deaths in Texas. Over the past century, global flood deaths declined by more than 80 percent. That happened not because nations reduced rainfall but because they have learned how to live with it. They built levees, dams, and drainage systems. They developed early warning systems and evacuated people before the water arrived. Kerr County in Texas failed to do any of these things. Despite its location in one of the most flood-prone regions in the United States, the county had no formal flood warning system in place. There were no sirens, no automated text alerts, no rapid evacuation protocol. The river rose, and families had no idea it was coming... Please subscribe now to watch the full video, read the full article, and support Public's award-winning reporting!

Michael Shellenberger

69,644 views • 1 year ago

The Leona River is on the rise in Uvalde, and the water is still climbing. Mandatory evacuations have already been ordered in some areas as floodwaters continue to surge. This isn’t just a Leona River story. Rivers across South Texas, including the Nueces, Frio, Sabinal, and Oasis, are surging toward major flood stage. Some forecast crests could challenge or exceed records, and the National Weather Service warns the flooding is far from over. For many communities across this part of Texas, these are the highest water levels they’ve seen since the devastating floods of the 1990s, and in some places, not since the late 1970s. It’s another reminder that in Texas’ Flash Flood Alley, rivers can transform from quiet streams into life-threatening torrents with astonishing speed. Stay with The Weather Channel for continuing coverage as this dangerous flood event unfolds.
1:24

The Leona River is on the rise in Uvalde, and the water is still climbing. Mandatory evacuations have already been ordered in some areas as floodwaters continue to surge. This isn’t just a Leona River story. Rivers across South Texas, including the Nueces, Frio, Sabinal, and Oasis, are surging toward major flood stage. Some forecast crests could challenge or exceed records, and the National Weather Service warns the flooding is far from over. For many communities across this part of Texas, these are the highest water levels they’ve seen since the devastating floods of the 1990s, and in some places, not since the late 1970s. It’s another reminder that in Texas’ Flash Flood Alley, rivers can transform from quiet streams into life-threatening torrents with astonishing speed. Stay with The Weather Channel for continuing coverage as this dangerous flood event unfolds.

Justin Michaels

16,687 views • 27 days ago

🚨#BREAKING: A life threatening flash flood emergency has been declared with multiple fatalities reported 📌#Kerrville | #Texas At this time, the National Weather Service in Texas has issued a flash flood emergency as life-threatening, catastrophic flooding is occurring in Kerrville Texas. Bridges, roadways, and multiple homes have been washed away, along with numerous vehicles. Water rescues are currently underway as rapidly rising river levels have even swept away a large stage in a park that was set to be used for Fourth of July celebrations. At least six fatalities have been confirmed so far, and more rain is expected. Police are urging all residents near the water to evacuate immediately.
0:36

🚨#BREAKING: A life threatening flash flood emergency has been declared with multiple fatalities reported 📌#Kerrville | #Texas At this time, the National Weather Service in Texas has issued a flash flood emergency as life-threatening, catastrophic flooding is occurring in Kerrville Texas. Bridges, roadways, and multiple homes have been washed away, along with numerous vehicles. Water rescues are currently underway as rapidly rising river levels have even swept away a large stage in a park that was set to be used for Fourth of July celebrations. At least six fatalities have been confirmed so far, and more rain is expected. Police are urging all residents near the water to evacuate immediately.

R A W S A L E R T S

686,835 views • 1 year ago

Someone just stole from 37,000 $TAO holders and walked away clean. Not because they broke a rule. Because no rule existed to stop them. That changes today. Here is what happened. Covenant AI ran one of the most watched subnets on Bittensor. On April 10, the founder sold their entire position and disappeared. No warning. No announcement. No on-chain signal. By the time holders found out, the price had already moved against them. This was not a hack. This was not a bug. This was a founder legally exiting into their own community with zero accountability. Bittensor just closed that door permanently. The Conviction upgrade is live on mainnet today. Every emission a subnet owner earns now locks automatically the moment it arrives. They cannot touch it immediately. If they want to exit they must submit a public unlock transaction on-chain. Visible to every single person on the network the second it is submitted. Then the clock starts. 30 days before 63% of their position becomes spendable. 90 days before 95% is accessible. You now have a month of warning before the first dollar of sell pressure hits. A silent exit is no longer possible. The founder has to tell you they are leaving before they can leave. And it goes further. Any holder can now lock their tokens toward a different address they believe would run the subnet better. The address with the most locked support behind it becomes eligible to take over entirely. Bad owners can be replaced by the community before they do damage. Before today, subnet investing had one risk nobody could price. The person running it could vanish overnight, and you would never see it coming. That risk has been removed from the equation. Skin in the game used to be a promise. Now, it is a number on a block explorer that every holder can verify in real time. The people who understand what accountable infrastructure means for the value of $TAO will not need to explain themselves later. This is still early.
23:08

Someone just stole from 37,000 $TAO holders and walked away clean. Not because they broke a rule. Because no rule existed to stop them. That changes today. Here is what happened. Covenant AI ran one of the most watched subnets on Bittensor. On April 10, the founder sold their entire position and disappeared. No warning. No announcement. No on-chain signal. By the time holders found out, the price had already moved against them. This was not a hack. This was not a bug. This was a founder legally exiting into their own community with zero accountability. Bittensor just closed that door permanently. The Conviction upgrade is live on mainnet today. Every emission a subnet owner earns now locks automatically the moment it arrives. They cannot touch it immediately. If they want to exit they must submit a public unlock transaction on-chain. Visible to every single person on the network the second it is submitted. Then the clock starts. 30 days before 63% of their position becomes spendable. 90 days before 95% is accessible. You now have a month of warning before the first dollar of sell pressure hits. A silent exit is no longer possible. The founder has to tell you they are leaving before they can leave. And it goes further. Any holder can now lock their tokens toward a different address they believe would run the subnet better. The address with the most locked support behind it becomes eligible to take over entirely. Bad owners can be replaced by the community before they do damage. Before today, subnet investing had one risk nobody could price. The person running it could vanish overnight, and you would never see it coming. That risk has been removed from the equation. Skin in the game used to be a promise. Now, it is a number on a block explorer that every holder can verify in real time. The people who understand what accountable infrastructure means for the value of $TAO will not need to explain themselves later. This is still early.

2xnmore

19,528 views • 3 months ago

Q: Here in Monaco, a driver can make a difference. Is it true? And how can a driver do that difference? "I mean, Monaco is one of the ones that the driver probably has more because it's a street circuit, a lot of risk that's taken, but I think you still need a good car. I've never driven this car at this track, so that will be new for me. It's a new characteristic to what I've experienced in the past, but it's exciting to drive for the first time a Ferrari around these streets of Monte Carlo."
0:29

Q: Here in Monaco, a driver can make a difference. Is it true? And how can a driver do that difference? "I mean, Monaco is one of the ones that the driver probably has more because it's a street circuit, a lot of risk that's taken, but I think you still need a good car. I've never driven this car at this track, so that will be new for me. It's a new characteristic to what I've experienced in the past, but it's exciting to drive for the first time a Ferrari around these streets of Monte Carlo."

sim

132,705 views • 1 year ago

This is very true. AI integration is both hard and necessary for companies. And AI-native junior employees can drive this adoption. Earlier this year Aaron Levie explained with martin_casado on the a16z podcast how AI fluency is one of the most powerful skills a new grad can have. The real value comes from knowing how to make AI work inside an organization. Most companies, both large and small, are slow to adopt new technology. They need people who can explain what AI can do, show them how to use it, and help them move faster. If you are fluent in AI, you can be that person. You can walk into a company, spot the areas where AI can save time or create new opportunities, and make it happen. That is a wedge into your first job and a way to stand out from everyone else.
0:39

This is very true. AI integration is both hard and necessary for companies. And AI-native junior employees can drive this adoption. Earlier this year Aaron Levie explained with martin_casado on the a16z podcast how AI fluency is one of the most powerful skills a new grad can have. The real value comes from knowing how to make AI work inside an organization. Most companies, both large and small, are slow to adopt new technology. They need people who can explain what AI can do, show them how to use it, and help them move faster. If you are fluent in AI, you can be that person. You can walk into a company, spot the areas where AI can save time or create new opportunities, and make it happen. That is a wedge into your first job and a way to stand out from everyone else.

a16z

366,027 views • 11 months ago

🚨#BREAKING: Over 50 hospital staff workers along with patients are stranded on top of Hospital roof due to life-threatening flood emergency 📌#Erwin | #Tennessee Currently, more than 50 people are stranded on the roof of Unicoi County Hospital in Erwin, Tennessee, after emergency crews attempted to evacuate people by ambulance and boats. Following a flash flood emergency that was declared from the national weather service due to heavy rainfall, the situation worsened as the water rose far too quickly and began moving with dangerous speed, making it impossible to safely extract those inside. Forced to seek refuge on the roof, patients and staff now await rescue. The National Guard has been called in to assist and rescue this is still developing
0:51

🚨#BREAKING: Over 50 hospital staff workers along with patients are stranded on top of Hospital roof due to life-threatening flood emergency 📌#Erwin | #Tennessee Currently, more than 50 people are stranded on the roof of Unicoi County Hospital in Erwin, Tennessee, after emergency crews attempted to evacuate people by ambulance and boats. Following a flash flood emergency that was declared from the national weather service due to heavy rainfall, the situation worsened as the water rose far too quickly and began moving with dangerous speed, making it impossible to safely extract those inside. Forced to seek refuge on the roof, patients and staff now await rescue. The National Guard has been called in to assist and rescue this is still developing

R A W S A L E R T S

3,025,281 views • 1 year ago

Journalist Alex Newman: "[AI] is a tool, and it can be used for good purposes, and it can be used for evil purposes." "When you look at the people who are coding this, many of them have documented incredibly nefarious intentions. They believe there's too many of us on the planet. They believe that we're better off without individual rights, that we're better off being ruled by them." "So when you take a tool that powerful and you put it in the hands of people that evil, you're asking for trouble." "We're talking about something here that can not just watch and surveil everything you do, but that also can start manipulating the things that you do." "This is an incredible time, and a time filled with incredible danger." Credit: Alex Newman Jefferey Jaxen The HighWire
2:38

Journalist Alex Newman: "[AI] is a tool, and it can be used for good purposes, and it can be used for evil purposes." "When you look at the people who are coding this, many of them have documented incredibly nefarious intentions. They believe there's too many of us on the planet. They believe that we're better off without individual rights, that we're better off being ruled by them." "So when you take a tool that powerful and you put it in the hands of people that evil, you're asking for trouble." "We're talking about something here that can not just watch and surveil everything you do, but that also can start manipulating the things that you do." "This is an incredible time, and a time filled with incredible danger." Credit: Alex Newman Jefferey Jaxen The HighWire

Wide Awake Media

19,864 views • 1 year ago

All I can think as I see the results roll in from Hungary is that a lot of people in the ‘western’ media owe Orbán Viktor an apology. They tried to destroy not only him, but his country, over hysterical claims that Orban’s Fidesz was authoritarian and rigged the system so this could NEVER happen. Orban saved his country from complete destruction—more than once. How many western leaders can actually say that? None of them. It’s why they hate him.
3:12

All I can think as I see the results roll in from Hungary is that a lot of people in the ‘western’ media owe Orbán Viktor an apology. They tried to destroy not only him, but his country, over hysterical claims that Orban’s Fidesz was authoritarian and rigged the system so this could NEVER happen. Orban saved his country from complete destruction—more than once. How many western leaders can actually say that? None of them. It’s why they hate him.

Bradley Devlin

20,875 views • 4 months ago

“The average person lives a life of luxury that previous generations could never imagine, but because of that we dont actually taste this deeper joy that I think God intends for us.” Most people today live with a level of comfort that would have looked unbelievable to almost anyone in history. Food is always available, entertainment never stops, boredom can be eliminated instantly, and almost every inconvenience can be solved with a screen and a few clicks. That’s why so many people today have everything they thought they wanted and still feel empty afterward. The strange paradox of modern life is that we have removed many of the hardships that once burdened humanity, but in the process we also removed many of the conditions that produced gratitude, resilience, depth, and dependence on God. A civilization drowning in comfort often becomes spiritually numb without even realizing it. Which is probably why some of the holiest people in history voluntarily embraced hardship even when they did not have to. They understood something we are slowly forgetting: Comfort and happiness are not the same thing.
1:09

“The average person lives a life of luxury that previous generations could never imagine, but because of that we dont actually taste this deeper joy that I think God intends for us.” Most people today live with a level of comfort that would have looked unbelievable to almost anyone in history. Food is always available, entertainment never stops, boredom can be eliminated instantly, and almost every inconvenience can be solved with a screen and a few clicks. That’s why so many people today have everything they thought they wanted and still feel empty afterward. The strange paradox of modern life is that we have removed many of the hardships that once burdened humanity, but in the process we also removed many of the conditions that produced gratitude, resilience, depth, and dependence on God. A civilization drowning in comfort often becomes spiritually numb without even realizing it. Which is probably why some of the holiest people in history voluntarily embraced hardship even when they did not have to. They understood something we are slowly forgetting: Comfort and happiness are not the same thing.

Athenaeum Book Club

64,373 views • 3 months ago

Replit CEO Amjad Masad on why the most ambitious employees are no longer blocked by engineering: "How can I upgrade my workforce for them to become generalist business people that can wield AI for the benefit of our customers and our bottom line?" "The most ambitious people are creating billions of dollars of value of their company." "There are people that are closest to the customer that feel that they have ideas that could make more revenue for the business, but they're often blocked by engineering." "Now a lot of them are bringing in Replit to work. They're building that idea, and they're making the money for their company. And then they're getting promoted, and then given more power." "I'm going to build a team of vibe coders that are going to go around the company and find all the inefficiencies and go solve them." "So we have a new role of this generalist automator." Amjad Masad with
1:40

Replit CEO Amjad Masad on why the most ambitious employees are no longer blocked by engineering: "How can I upgrade my workforce for them to become generalist business people that can wield AI for the benefit of our customers and our bottom line?" "The most ambitious people are creating billions of dollars of value of their company." "There are people that are closest to the customer that feel that they have ideas that could make more revenue for the business, but they're often blocked by engineering." "Now a lot of them are bringing in Replit to work. They're building that idea, and they're making the money for their company. And then they're getting promoted, and then given more power." "I'm going to build a team of vibe coders that are going to go around the company and find all the inefficiencies and go solve them." "So we have a new role of this generalist automator." Amjad Masad with

a16z

94,601 views • 5 months ago

Sam Altman says the four hour work week isn't coming, even with AI: "Technology for a long time has been promising people that they’re going to work less and have all this leisure. And it has gone in that direction." "But somehow we never get the promise of the four hour workweek at mass scale in society. And I don’t expect AI to change that." "Clearly, society is not working for a lot of people, and more productivity gains that actually accrue to people would be a great thing. And I suspect they will. I suspect that will happen."
0:44

Sam Altman says the four hour work week isn't coming, even with AI: "Technology for a long time has been promising people that they’re going to work less and have all this leisure. And it has gone in that direction." "But somehow we never get the promise of the four hour workweek at mass scale in society. And I don’t expect AI to change that." "Clearly, society is not working for a lot of people, and more productivity gains that actually accrue to people would be a great thing. And I suspect they will. I suspect that will happen."

dnap

195,667 views • 17 days ago

Chamath: “Private equity in general is totally hosed.” 🏢🚨 “I think the history of this is important.” “There was a long standing belief that the best way to generate the best risk adjusted return was to have what's called a 60/40 allocation. 60% to bonds and 40% to equities.” “Over many years, especially when we artificially suppressed rates at zero, a lot of people started to move their allocations away from 60/40 and they started to make more and more investments further out on the risk curve.” “The biggest beneficiaries of that were venture capital, private equity, and hedge funds.” “The thing with private equity is that because rates were zero, they had an infinite amount of borrowing capacity at very little downside to them, and so they were able to manufacture returns much faster than venture capital and hedge funds could.” “So as a result, you had an initial group of people that were defining the asset class, making a ton of money, and then you had all these fast followers that said, ‘Well, if they're doing it, I can do it too.’” “But then always what happens is then you have this flood of laggards that just flood the zone.” “And it's these laggards that make it very difficult to generate returns because they start overpaying for assets, they start mismanaging and under managing the assets that they do own.” “That created a lot of competition, and so that's why you see this hockey stick graph.” “And when you see that kind of graph, it doesn't matter what asset class it is. The returns go to zero.” “And so we've seen this in venture capital. We've seen this in hedge funds. And we're now going to see this in private equity.”
1:36

Chamath: “Private equity in general is totally hosed.” 🏢🚨 “I think the history of this is important.” “There was a long standing belief that the best way to generate the best risk adjusted return was to have what's called a 60/40 allocation. 60% to bonds and 40% to equities.” “Over many years, especially when we artificially suppressed rates at zero, a lot of people started to move their allocations away from 60/40 and they started to make more and more investments further out on the risk curve.” “The biggest beneficiaries of that were venture capital, private equity, and hedge funds.” “The thing with private equity is that because rates were zero, they had an infinite amount of borrowing capacity at very little downside to them, and so they were able to manufacture returns much faster than venture capital and hedge funds could.” “So as a result, you had an initial group of people that were defining the asset class, making a ton of money, and then you had all these fast followers that said, ‘Well, if they're doing it, I can do it too.’” “But then always what happens is then you have this flood of laggards that just flood the zone.” “And it's these laggards that make it very difficult to generate returns because they start overpaying for assets, they start mismanaging and under managing the assets that they do own.” “That created a lot of competition, and so that's why you see this hockey stick graph.” “And when you see that kind of graph, it doesn't matter what asset class it is. The returns go to zero.” “And so we've seen this in venture capital. We've seen this in hedge funds. And we're now going to see this in private equity.”

The All-In Podcast

800,205 views • 10 months ago

🐹 also, this might be a bit of a deep talk (topic). actually, there really isn't much in this world that lasts forever. but still, i hope that villains and xdiz can be one of those rare exceptions. lately, i've found myself thinking that more and more. 🐹 right now, xdiz is receiving so much more love than before. more people are discovering us, we have so many new fans, and every year there are even more people celebrating my birthday than the year before. i never want to take any of that for granted. 🐹 the love you've all given us... i want to work even harder so i can give it back to you. i'll keep doing my best to become a leader that villains can be proud of.
0:52

🐹 also, this might be a bit of a deep talk (topic). actually, there really isn't much in this world that lasts forever. but still, i hope that villains and xdiz can be one of those rare exceptions. lately, i've found myself thinking that more and more. 🐹 right now, xdiz is receiving so much more love than before. more people are discovering us, we have so many new fans, and every year there are even more people celebrating my birthday than the year before. i never want to take any of that for granted. 🐹 the love you've all given us... i want to work even harder so i can give it back to you. i'll keep doing my best to become a leader that villains can be proud of.

ten 🪐

11,356 views • 20 days ago

*New Paper on AI & Democracy* Imagine two approaches to democracy. The one we have today, where citizens choose a professional politician to represent them and others. Or an augmented form of democracy, where each citizen controls a personalized AI that helps them participate in thousands of nuanced decisions. This second approach is the idea of Augmented Democracy I introduced six years ago at TED. In our latest paper we explore a simplified version of Augmented Democracy by combining off-the-shelf LLMs, such as ChatGPT, with data collected using a collaborative government program builder. This was an online game where people build a personalized government program using proposals extracted from the programs of the candidates of the 2022 presidential election in Brazil. So how accurate are these augmented forms of democracy? Imagine a user who gave us 40 answers. We can use the first 20 to fine-tune a model that we can test using the 20 answers the model didn’t see. We can then compare the accuracy of these predictions with the ones obtained by a “bundle” rule, which assumes that users that self-reported to be from the left or right always chose the proposals from the candidate that shares their political identity. This showed us that LLMs were more accurate at predicting policy preferences than the bundle rule, meaning that the preferences captured in the participation data were more nuanced than a left-right axis, and that the LLMs can capture some of that nuance. Also, the LLMs can choose among policies coming from the same candidate, which is something that we cannot do using a bundle rule. But can these LLMs help us complete the aggregate preferences of the population? Direct or unbundled forms of participation can result in incomplete data when people answer only a fraction of all questions. In our paper, we simulate this incompleteness by sampling the full dataset. We ask how close we can get to the full dataset by using a random sample, or a random sample augmented by predictions made by these LLMs. Overall, we find that LLM-augmented data gets much closer to the full dataset than a pure random sample. These results do not mean that augmented democracy technology is ready, but they means we are in a much better place to continue exploring this idea than six years ago. This paper was a collaborative effort with Jairo Gudino, PhD student at CCL at the University of Toulouse Capitole and Umberto Grandi from IRIT also at the University of Toulouse Capitole. We hope you find these results insightful!
2:16

*New Paper on AI & Democracy* Imagine two approaches to democracy. The one we have today, where citizens choose a professional politician to represent them and others. Or an augmented form of democracy, where each citizen controls a personalized AI that helps them participate in thousands of nuanced decisions. This second approach is the idea of Augmented Democracy I introduced six years ago at TED. In our latest paper we explore a simplified version of Augmented Democracy by combining off-the-shelf LLMs, such as ChatGPT, with data collected using a collaborative government program builder. This was an online game where people build a personalized government program using proposals extracted from the programs of the candidates of the 2022 presidential election in Brazil. So how accurate are these augmented forms of democracy? Imagine a user who gave us 40 answers. We can use the first 20 to fine-tune a model that we can test using the 20 answers the model didn’t see. We can then compare the accuracy of these predictions with the ones obtained by a “bundle” rule, which assumes that users that self-reported to be from the left or right always chose the proposals from the candidate that shares their political identity. This showed us that LLMs were more accurate at predicting policy preferences than the bundle rule, meaning that the preferences captured in the participation data were more nuanced than a left-right axis, and that the LLMs can capture some of that nuance. Also, the LLMs can choose among policies coming from the same candidate, which is something that we cannot do using a bundle rule. But can these LLMs help us complete the aggregate preferences of the population? Direct or unbundled forms of participation can result in incomplete data when people answer only a fraction of all questions. In our paper, we simulate this incompleteness by sampling the full dataset. We ask how close we can get to the full dataset by using a random sample, or a random sample augmented by predictions made by these LLMs. Overall, we find that LLM-augmented data gets much closer to the full dataset than a pure random sample. These results do not mean that augmented democracy technology is ready, but they means we are in a much better place to continue exploring this idea than six years ago. This paper was a collaborative effort with Jairo Gudino, PhD student at CCL at the University of Toulouse Capitole and Umberto Grandi from IRIT also at the University of Toulouse Capitole. We hope you find these results insightful!

César A. Hidalgo

26,915 views • 1 year ago

Meta's new AI can predict your brain better than a brain scan. TRIBE v2 is a foundation model trained on 1,000+ hours of brain imaging data from 720 people. You feed it a video, sound clip, or text, and it predicts: > Which brain regions light up > How strongly > And in what order When tested on people it had never seen, the model's predictions were actually more accurate than most real brain scans (which get distorted by heartbeats, breathing, and movement). Researchers then replicated decades of classic neuroscience experiments entirely inside the software. No scanner, no human subjects. The model correctly identified the brain's face recognition center, language network, and emotional processing regions on its own. But look at what else Meta has been building: > Ray-Ban smart glasses that see and hear what you do > A wristband that reads nerve signals > And now a model that predicts how your brain responds to any piece of content There's no evidence these are all connected... but Meta now has a complete picture of attention, from the stimulus to the neural response.
1:08

Meta's new AI can predict your brain better than a brain scan. TRIBE v2 is a foundation model trained on 1,000+ hours of brain imaging data from 720 people. You feed it a video, sound clip, or text, and it predicts: > Which brain regions light up > How strongly > And in what order When tested on people it had never seen, the model's predictions were actually more accurate than most real brain scans (which get distorted by heartbeats, breathing, and movement). Researchers then replicated decades of classic neuroscience experiments entirely inside the software. No scanner, no human subjects. The model correctly identified the brain's face recognition center, language network, and emotional processing regions on its own. But look at what else Meta has been building: > Ray-Ban smart glasses that see and hear what you do > A wristband that reads nerve signals > And now a model that predicts how your brain responds to any piece of content There's no evidence these are all connected... but Meta now has a complete picture of attention, from the stimulus to the neural response.

Rowan Cheung

29,939 views • 4 months ago

What does a rifle using subsonic ammo sound like? What type of muzzle flash can you expect? Is it so quiet you can't hear it? Here's a video I shot over a year ago when people were saying there was "muzzle flash" from the AGR building in Butler. Spoiler Alert: While at the time I could not definitively say what it was (it was later confirmed as a reflection of headlights from a police cruiser), I knew immediately, it was NOT muzzle flash. This is .223 with a very short barrel which is the most likely to produce muzzle flash. 300BLK or 8.6BLK or just about any other caliber will be slightly quieter and produce even less muzzle flash. Takeaway: If you don't see muzzle flash from this, you won't see muzzle flash from any other rifle that is suppressed and using subsonic ammo.
4:10

What does a rifle using subsonic ammo sound like? What type of muzzle flash can you expect? Is it so quiet you can't hear it? Here's a video I shot over a year ago when people were saying there was "muzzle flash" from the AGR building in Butler. Spoiler Alert: While at the time I could not definitively say what it was (it was later confirmed as a reflection of headlights from a police cruiser), I knew immediately, it was NOT muzzle flash. This is .223 with a very short barrel which is the most likely to produce muzzle flash. 300BLK or 8.6BLK or just about any other caliber will be slightly quieter and produce even less muzzle flash. Takeaway: If you don't see muzzle flash from this, you won't see muzzle flash from any other rifle that is suppressed and using subsonic ammo.

Paramount Tactical - Gary Melton

97,360 views • 10 months ago

Never have I ever seen this in my 10 years of bl. Almost 12 hours spanning two days of concert performance, with numerous special guests, performances, set design, costume fitting, choreography, acting, and a star vocals. Not to mention this was at the IMPACT Arena which housed almost a total of 24,000 people in 2 days. One of the most influential venues that have housed many global stars previously. Not even kpop concerts have that must production, quality as much as what I have witnessed in the last two days. Khemjira showed the world that you can have the entire cast be star actors and be outstanding performers. Also having deeper meaning in Thai representation. Khemjira is more than a show. It's an experience for a life time you can never forget.
0:20

Never have I ever seen this in my 10 years of bl. Almost 12 hours spanning two days of concert performance, with numerous special guests, performances, set design, costume fitting, choreography, acting, and a star vocals. Not to mention this was at the IMPACT Arena which housed almost a total of 24,000 people in 2 days. One of the most influential venues that have housed many global stars previously. Not even kpop concerts have that must production, quality as much as what I have witnessed in the last two days. Khemjira showed the world that you can have the entire cast be star actors and be outstanding performers. Also having deeper meaning in Thai representation. Khemjira is more than a show. It's an experience for a life time you can never forget.

dia

50,983 views • 6 months ago

Johns Hopkins University has been a WHO collaborating center for vaccine epidemiology, with the terms of that agreement explicitly tied to supporting the WHO's Immunization 2030 agenda, a program calling for vaccines at every stage of life from birth to death, with 500 vaccines currently in development or planned. Dr. Tess Lawrie has been warning about this trajectory for years. The framework is already written. The institutions behind it are the most credentialed names in global public health. And the hantavirus coverage these past few weeks is a live demonstration of how the pipeline works: a disease that kills a handful of people annually generates national fear, and a vaccine partnership launches before the news cycle ends. If seven people die from a pathogen but seven billion can be persuaded to inject themselves with it, the math works out very well for the people running the program. The fear is the mechanism. The injection is the destination.
1:24

Johns Hopkins University has been a WHO collaborating center for vaccine epidemiology, with the terms of that agreement explicitly tied to supporting the WHO's Immunization 2030 agenda, a program calling for vaccines at every stage of life from birth to death, with 500 vaccines currently in development or planned. Dr. Tess Lawrie has been warning about this trajectory for years. The framework is already written. The institutions behind it are the most credentialed names in global public health. And the hantavirus coverage these past few weeks is a live demonstration of how the pipeline works: a disease that kills a handful of people annually generates national fear, and a vaccine partnership launches before the news cycle ends. If seven people die from a pathogen but seven billion can be persuaded to inject themselves with it, the math works out very well for the people running the program. The fear is the mechanism. The injection is the destination.

The HighWire

19,498 views • 2 months ago