Report warns AI could overwhelm system made to curb online child exploitation
A new report from the Stanford Internet Observatory revealed that the CyberTipline for reporting online child exploitation (CSAM) is overwhelmed and needs improvement for law enforcement to capture predators. The CyberTipline is “enormously valuable and leads to the rescue of children and prosecution of offenders,” however, the report said child sex abuse material created by artificial intelligence is threatening to overwhelm the situation.
The system is currently bogged down with millions of tips and law enforcement officers are not able to prioritize the reports for investigation.
The CyberTipline was created by the National Center for Missing and Exploited Children (NCMEC) in 1998 and “established a national resource center and clearinghouse to provide technical assistance to state and local governments, law enforcement agencies and individuals in locating and recovering missing children.”
The CyberTipline allows the public and electronic service providers to report online child sex abuse exploitation. Staffers working the CyberTipline review the reports, identify the location of the victim or predator and send the information to the appropriate local law enforcement agency.
If a report cannot be geolocated to a specific area, the report is then made available to federal law enforcement.
In 2023, the center received more than 36 million reports of child exploitation. Of those, 2.1 million CyberTipline reports were given to federal law enforcement agencies.
“Almost certainly in the years to come, the CyberTipline will just be flooded with highly realistic-looking AI content, which is going to make it even harder for law enforcement to identify real children who need to be rescued,” researcher Shelby Grossman said.
Just 5%-8% of the tips lead to arrests because of a lack of funding and legalities, according to the report.
In 2023, about 92% of the tips received involved countries outside of the U.S. — a stark difference from 2008 when the majority of the tips concerned American victims and offenders.
The Stanford report outlines a few recommendations such as having Congress increase funding to NCMEC, providing clear information to social media platforms on how to file a detailed report to the CyberTipline, and providing resources to law enforcement so they have the time and investigate child sex abuse crimes.
Several AI companies — including OpenAI, Meta and Google — have agreed to implement new practices and principles to minimize the risk of child sexual abuse material and train to remove the material from their platforms.
Weather forecasting breakthrough: AI may save lives with new technology
Google unveiled a potentially groundbreaking development in weather forecasting. The company’s latest innovation, known as its “Scalable Ensemble Diffusion Sample” (SEEDS), is using cutting-edge artificial intelligence to revolutionize weather forecasting.
Similar to language models like ChatGPT, SEEDS taps into the power of AI to churn out forecasts in record time. SEEDS is different due to its ability to dig deeper into weather extremes. The technology can detect dangerous events that may slip through the cracks of traditional weather models.
Scientists say that with climate change, extreme weather events are becoming more frequent and making accurate and timely forecasts is more vital than ever before.
SEEDS is also cheaper than traditional weather models. While traditional models use a wide range of forecasts to predict 10-50 weather scenarios, SEEDS needs just one or two forecasts to predict 31 different weather scenarios.
Another difference between SEEDS and traditional weather models is that traditional models use a set of initial conditions and random variables to generate forecasts, like humidity, temperature, wind speed and air pressure. However, traditional methods often struggle to accurately predict extreme weather as atmospheric dynamics change.
In contrast, SEEDS approach focuses on the relationships between different weather variables such as energy and sea level pressure. The approach allows for more detailed forecast models, providing a more thorough picture of future weather conditions.
For example, SEEDS played a crucial role in predicting Europe’s 2022 heatwave. Just seven days before the heatwave hit, traditional forecasting methods showed no indication of the extreme temperatures. However, SEEDS predicted the event, providing advanced warnings and allowing people to plan, possibly saving lives.
AI agents go nose-to-nose with human pilots in real-world dogfight
For the first time in the history of jets, artificial intelligence and humans, an AI pilot flew a fighter jet head-to-head against a human pilot in the real world. The test took place sometime in 2023.
Pulling off the feat off was a collaborative effort between several government agencies including the Air Force Research Lab, the Defense Advanced Research Projects Agency (DARPA), the Air Force Test Center, and the Air Force Test Pilot School. Hitting the milestone also required help from industry partners like Lockheed Martin, EpiSci and Shield AI.
The teams worked together to put AI agents at the controls of the X-62A Vista. The plane is a test aircraft based on an F-16 airframe.
“The X-62A is an incredible platform, not just for research and advancing the state of tests, but also for preparing the next generation of test leaders,” Col. James Valpiani, commandant of the Test Pilot School, said. “When ensuring the capability in front of them is safe, efficient, effective and responsible, industry can look to the results of what the X-62A [Air Combat Evolution] team has done as a paradigm shift. We’ve fundamentally changed the conversation by showing this can be executed safely and responsibly.”
In less than a year, the project’s partners went from installing live AI agents into the X-62A, to actually handing over the controls in the first human vs. AI dogfights. More than 100,000 lines of code in the flight-critical software were updated during testing.
It took time to teach the AI agents the rules of the sky and how to operate within the domain under real-world conditions. Over the course of 21 test flights, the machine-based learning tools were taught how to perform defensive maneuvers before going into the more advanced high-aspect, nose-to-nose engagements.
During the dogfights, human pilots came within 2,000 feet of the AI-piloted Vista while moving at around 1,200 miles per hour.
“The potential for autonomous air-to-air combat has been imaginable for decades, but the reality has remained a distant dream up until now,” Air Force Secretary Frank Kendall said. “In 2023, the X-62A broke one of the most significant barriers in combat aviation. This is a transformational moment, all made possible by breakthrough accomplishments of the X-62A [Air Combat Evolution] team.”
Secretary Kendall will get a front seat to the new tech when he takes a ride in the AI-piloted plane later this year.
While teaching an artificial intelligence to dogfight, and actually having that intelligence perform the maneuvers in the real-world is absolutely monumental, the Air Force’s chief test Pilot said focusing solely on that aspect misses the point of the overall moment.
“Dogfighting was the problem to solve so we could start testing autonomous artificial intelligence systems in the air,” Bill Gray said. “Every lesson we’re learning applies to every task you could give to an autonomous system.”
Neither the AFRL nor DARPA revealed who won the dogfights between human test pilots and their artificial counterparts. However, when AI agents went up against humans in a simulated environment, the AI agents swept the competition.
Meta AI debuts across platforms raising privacy, bias concerns
After months of anticipation, Meta launched the company’s AI chatbot, Meta AI, on Thursday, April 18. Meta AI integrates advanced AI technology across everyday applications such as Facebook, Instagram, Messenger, WhatsApp and Ray-Ban smart glasses. The rollout is sparking concerns about the technology’s controversial impact on digital interactions.
Like its generative AI competitors, Meta AI can create images, craft captions and even act as a moderator in group chats, addressing questions when needed.
Anika Navaroli, a senior fellow at Columbia’s Tow Center for Digital Journalism, cautioned that AI tools may perpetuate cycles of misinformation and bias on social platforms, exacerbating societal divisions. She described these biases as creating a “feedback loop to hell,” which intensifies hate speech and harmful content. The feedback loops make biases increasingly difficult to manage or eliminate.
The company integrated Meta AI into almost all of the company’s platforms without an option to disable it. When users inquire about how to remove it, the chatbot provides instructions, but they are misleading and not based on accurate information.
Straight Arrow News
Users can avoid using Meta AI by ignoring its suggestions and prompts, but it will remain a prominent feature on the platform.
China’s AI-generated content targets US, India and South Korea elections
China and North Korea are intensifying their cyber operations, leveraging artificial intelligence to influence global opinions and elections in the U.S., South Korea, India, and Taiwan, according to Microsoft’s latest Threat Intelligence report. The report details China’s use of AI-generated content, such as videos, memes and fake news anchors, to sway public opinion and influence the outcomes of key elections.
Clint Watts of Microsoft’s Threat Analysis Center highlights China’s use of fake social media accounts to probe divisive U.S. domestic issues, aiming to understand what divides U.S. voters. These influence operations have targeted various issues within the U.S., gathering intelligence on American political views.
Last year, the group Storm 1376 falsely claimed the Maui fires were set by the U.S. government as a test of a “weather weapon.” The challenge lies in combating misinformation from realistic generative AI, as many refuse to accept such content as false, especially when it aligns with their beliefs and values.
In Taiwan, a suspected AI-generated audio falsely portrayed Foxconn’s Terry Gou endorsing another presidential candidate, which YouTube quickly removed. Meanwhile, North Korea has focused on cryptocurrency thefts and supply chain attacks, funding its military ambitions and enhancing intelligence collection through AI.
The U.N. estimates that since 2017, North Korean cyber actors have stolen over $3 billion in cryptocurrency. Microsoft acknowledges that while AI-generated content currently has a minimal impact on elections, China’s ongoing efforts to refine this content could become more effective as technology advances.
Jamie Dimon warns of higher interest rates, recession and AI taking jobs
In a year where Americans are anxiously waiting for interest rates to drop, the CEO of the world’s largest bank warns rates may go even higher than they are today and don’t count out a recession. There are two CEOs in this country whose annual letters to shareholders are more widely anticipated than the rest: Warren Buffett and Jamie Dimon.
JPMorgan Chase CEO Dimon’s April 8 letter pours water on investor enthusiasm that the Federal Reserve can orchestrate lower inflation while still growing the economy.
“These markets seem to be pricing in at a 70% to 80% chance of a soft landing…I believe the odds are a lot lower than that,” he wrote. “Therefore, we are prepared for a very broad range of interest rates, from 2% to 8% or even more, with equally wide-ranging economic outcomes — from strong economic growth with moderate inflation…to a recession with inflation; i.e., stagflation.”
Dimon said stagflation is the worst-case scenario.
“I love my country. Maybe one day I’ll serve my country in one capacity or another. But I love what I do,” Dimon told Bloomberg in May 2023.
Dimon’s annual letter comes at a time when people are floating his name as a possible pick for Treasury secretary should Donald Trump win the election. For years, people in political circles have questioned whether Dimon has political ambitions.
Dimon is not an ardent supporter of Trump and in November 2023, he encouraged people to back Trump’s opponent, Nikki Haley.
“Even if you’re a very liberal Democrat, I urge you, help Nikki Haley, too. Get a choice on the Republican side that might be better than Trump,” he said.
I love my country. Maybe one day I’ll serve my country in one capacity or another.
Jamie Dimon, JPMorgan Chase CEO
By January 2024, though, he was praising Trump’s policies, from NATO to immigration to economic growth. And in his letter advocating for U.S. support of Ukraine, he even italicized a Trump phrase, “America first.”
“My heart is Democratic but my brain is kind of Republican,” he said in 2019.
AI as transformational as the internet
When it came to listing the most important issues facing JPMorgan Chase, one easily rose to the top of the list: AI.
“We are completely convinced the consequences will be extraordinary,” Dimon wrote.
He said artificial intelligence could be as transformational as the printing press, the steam engine, electricity, computing and the internet.
Dimon said JPMorgan Chase currently employs more than 2,000 AI-machine learning experts and data scientists. And his predictions on how it will transform their workforce ripple throughout the economy.
“Over time, we anticipate that our use of AI has the potential to augment virtually every job, as well as impact our workforce composition. It may reduce certain job categories or roles, but it may create others as well,” he wrote.
He said affected employees would be retrained and redeployed, a task that’s also front of mind for tech giants. A group of them, including Google, Microsoft, IBM and Intel, are aiming to upskill or reskill 95 million workers in the next 10 years whose jobs AI threatens.
Billionaire Steve Cohen predicts work week will be reduced with help of AI
Steve Cohen, the billionaire owner of the Mets, predicts that the “four-day work week is coming,” thanks to advances in artificial intelligence. He shared his belief with CNBC’s “Squawk Box” on Wednesday, April 3.
Cohen didn’t claim to know when it would happen, however, he did say that it’s part of the reason he invested in TGL, a new golf startup league. He said innovations in the work week will lead to “a theme of more leisure for people, which means golf rounds will go up.”
Cohen said that the reduced work week is also due to the wide-held belief that work production dips on Fridays.
Recent advances in AI could also spur changes to the traditional work week. A 2023 study by Autonomy, a think tank group, showed gains in AI could reduce the work week to 32 hours per week for an estimated 28% of the workforces in the United States and United Kingdom by 2033. The data also showed there would be no drop-off in pay.
Researchers predicted that mundane tasks will eventually be done by AI, reducing the need for workers to clock in for 40 hours per week. Data indicated that some tasks could be achieved through ChatGPT and “Large Language Models.” The study found that by bringing Large Language Models on board, businesses could reduce the average employee’s hours by 10%.
Sen. Bernie Sanders, D-Vt., has previously advocated for a 32-hour work week. Sanders proposed a bill that would reduce the work week but would not reduce pay. However, Sanders legislation doesn’t appear to be gaining much traction.
Meanwhile, thousands of firms across the U.K. experimented with the world’s largest 32-hour work week trial in 2023. Of the companies who participated, 91% reportedly have kept the reduced hours.
A 32-hour work week pilot program in Spain also showed benefits to workers. Workers who participated in the pilot program reportedly felt less stress, were less tired and had more time for things they enjoy.
Workers who are employed by Cohen shouldn’t expect the change. Cohen said that as long as the market stays open on Friday, traders will be working. Still, Cohen believes “the vast majority” of workers are on track for a four-day work week.
Estate of George Carlin settles with podcast over AI-generated special
The estate of the late comedian George Carlin has settled a lawsuit with the “Dudesy” podcast, which had purportedly used an AI-generated version of Carlin to comment on current events in an audio special earlier this year. The alleged artificially generated comedy special imitating the late comedian was titled ‘I’m Glad I’m Dead.’
Filed in January, the lawsuit accused the podcast of infringing on the copyrights and publicity rights associated with Carlin’s name, image, and likeness. Carlin, who passed away in 2008, remains a significant figure in comedy.
Following the lawsuit, the podcast’s creators removed the special from their YouTube channel and, as part of the settlement, agreed not to re-upload the video or use Carlin’s image on any platform.
“I am pleased that this matter was resolved quickly and amicably, and I am grateful that the defendants acted responsibly by swiftly removing the video they made,” said Carlin’s daughter Kelly Carlin.
Carlin’s daughter emphasized its importance as a cautionary tale regarding the use of AI and the necessity for proper safeguards.
7.4 magnitude earthquake rocks Taiwan, several dead
Several people are dead in the strongest earthquake to hit Taiwan in 25 years, and the future of Disney will be decided today in a boardroom vote. These stories and more highlight The Morning Rundown for Wednesday, April 3, 2024.
Several dead after major earthquake rocks Taiwan
At least nine people have died and more than 800 are injured following the largest earthquake to strike Taiwan in 25 years, occurring Wednesday, April 3. The aftermath saw buildings tilting precariously, with rescuers employing ladders to reach those trapped inside. The event brought traffic to a halt and led to the suspension of train services throughout the island.
Get up to speed on the stories leading the day every weekday morning. Get The Morning RundownTM newsletter straight to your inbox!
Taiwan’s authorities recorded the earthquake at a magnitude of 7.2, whereas the U.S. Geological Survey reported it as 7.4. The earthquake hit around 8 a.m. local time, prompting tsunami warnings for southern Japan and the Philippines, which were later lifted.
Biden ‘outraged’ over Israeli airstrike that killed aid workers in Gaza
President Joe Biden said he was “outraged and heartbroken” by the deaths of seven World Central Kitchen aid workers in Gaza, caused by Israeli airstrikes on Monday, April 1. This marks his most forceful condemnation of Israel amid the ongoing conflict with Hamas.
Biden criticized Israel for not doing enough to safeguard humanitarian workers in the region. Following the incident, World Central Kitchen has temporarily halted its operations in Gaza.
Israeli Prime Minister Benjamin Netanyahu acknowledged that the strikes inadvertently harmed civilians and committed to a prompt investigation into the airstrikes targeting the vehicles of aid workers.
Trump, Biden win primaries but voters express concern
In the latest round of primaries, President Joe Biden and former President Donald Trump, who have secured their party’s nominations, continued to gather delegates. Despite the expected victories in Connecticut, New York, Rhode Island and Wisconsin, both encountered considerable dissent within their ranks.
Among Democrats, a significant segment of voters opted for “Uncommitted,” with protest votes hitting as high as 15 percent in Rhode Island.
On the Republican front, Nikki Haley, who withdrew from the race last month, still received support, securing at least 10 percent of the vote in all four states.
Results of Disney boardroom vote to be announced
The outcome of a prolonged and intense boardroom dispute concerning Disney’s future direction will be disclosed today during its annual shareholder meeting, marking the culmination of months of internal conflict.
In a significant moment of contention within the boardroom, stakeholders have been divided over maintaining the company’s current strategic path. Nelson Peltz has been vocal in his criticism, asserting that Disney has diverged from its core values, leading to a significant depreciation in its market value.
Peltz’s primary complaints center around what he perceives as mismanagement and strategic errors by the current leadership, as he seeks to convince shareholders of the need for a new direction.
According to Reuters’ latest report this morning, Disney is anticipated to have garnered sufficient support from its shareholders to maintain its course, with official confirmation expected later today.
Estate of George Carlin settles with podcast over AI-generated special
The estate of the late comedian George Carlin has settled a lawsuit with the “Dudesy” podcast, which had allegedly used an AI-generated version of Carlin to comment on current events in an audio special earlier this year.
Filed in January, the lawsuit accused the podcast of infringing on the copyrights and publicity rights associated with Carlin’s name, image, and likeness. Carlin, who passed away in 2008, remains a significant figure in comedy.
Following the lawsuit, the podcast’s creators removed the special from their YouTube channel and, as part of the settlement, agreed not to re-upload the video or use Carlin’s image on any platform.
Carlin’s daughter expressed her satisfaction with the swift resolution of the case and emphasized its importance as a cautionary tale regarding the use of AI and the necessity for proper safeguards.
Tropicana Las Vegas closes its doors after 67 years
Throughout its history, the Tropicana has played host to numerous legendary figures, including the Rat Pack, and was the backdrop for the debuts of magicians Siegfried and Roy. It was also featured in the 1971 James Bond film “Diamonds Are Forever,” starring Sean Connery.
The Tropicana expressed its gratitude to its employees, guests, and the city of Las Vegas for “67 years of unforgettable memories and unwavering support” in a farewell post on its official Instagram account.
Armed drone-on-drone violence documented in Ukraine
Time to add a couple more entries to the list of military records from the war in Ukraine. On March 29, video was captured showing what could be the first recorded combat between an armed unmanned aerial vehicle (UAV) and armed unmanned ground vehicle (UGV).
The video, posted to Telegram and analyzed by Army Recognition, documents an encounter south of Avdiivka between two Russian ground drones and Ukrainian FPV aerial drones.
The Russian UGVs appear to be immobilized while Ukrainian FPVs carrying explosive munitions hammer away, eventually destroying the Russian’s robotic devices.
The Russian UGVs appear to be armed with automatic grenade launchers. Army Recognition reported they were AGS-17s, a launcher developed during the Soviet era and known for its firepower.
The analysts at Army Recognition said the historical moment of the first conflict between armed ground and aerial drones showcases the flexibility the technology now brings to the battlefield, and further signifies the shift towards more autonomous forms of combat.
The Armed Forces of Ukraine are now flying FPV drones with auto-targeting capabilities.
In many of the FPV drone videos from Ukraine, the video feed goes to static before impact. That’s because the drone loses line-of-sight connection with its operators or the drone was successfully jammed with electronic countermeasures. Either way, the last part of the attack is more or less guessing where the target will be and hoping for a hit.
Now, video from Ukraine’s 60th and 63rd Mechanized Brigades appears to show an FPV drone still hitting its target long after the video feed dropped out. This indicates that the drone was guided to its final destination with the help of artificial intelligence, taking the human out of the loop.
While much of the rest of the world is still debating how drones and AI should be used in warfare, Ukraine and Russia are creating policy on the fly in real time.
The Ukraine war is the first major armed conflict in Europe in 25 years. Since the war broke out, the use of drones developed dramatically fast. Aerial drones now deal more damage for Ukraine than traditional artillery. Unmanned ground vehicles went from simple munitions mules and mine layers to full-on attack platforms and Ukraine’s Sea Baby drones forced Russia’s Black Sea fleet to flee the Black Sea.
None of this is going unnoticed. The United States and plenty of other militaries are adapting their own techniques and future weapons acquisitions based on the lessons learned in Ukraine.