AI understands many things... except for human social interactions
NEW YORK, April 27 — Artificial intelligence continues to advance, yet this technology still struggles to grasp the complexity of human interactions. A recent American study reveals that, while AI excels at recognising objects or faces in still images, it remains ineffective at describing and interpreting social interactions in a moving scene.
The team led by Leyla Isik, professor of cognitive science at Johns Hopkins University, investigated how artificial intelligence models understand social interactions. To do this, the researchers designed a large-scale experiment involving over 350 AI models specialising in video, image or language. These AI tools were exposed to short, three-second video sequences illustrating various social situations. At the same time, human participants were asked to rate the intensity of the interactions observed, according to several criteria, on a scale of 1 to 5. The aim was to compare human and AI interpretations, in order to identify differences in perception and better understand the current limits of algorithms in analysing our social behaviours.
'A blind spot in AI model development'
The human participants were remarkably consistent in their assessments, demonstrating a detailed and shared understanding of social interactions. AI, on the other hand, struggled to match these judgments. Models specialising in video proved particularly ineffective at accurately describing the scenes observed. Even models based on still images, although fed with several extracts from each video, struggled to determine whether the characters were communicating with each other. As for language models, they fared a little better, especially when given descriptions written by humans, but remained far from the level of performance of human observers.
For Leyla Isik, the inability of artificial intelligence models to understand human social dynamics is a major obstacle to their integration into real-world environments. 'AI for a self-driving car, for example, would need to recognise the intentions, goals, and actions of human drivers and pedestrians. You would want it to know which way a pedestrian is about to start walking, or whether two people are in conversation versus about to cross the street,' the study's lead author explains in a news release. 'Any time you want an AI to interact with humans, you want it to be able to recognise what people are doing. I think this [study] sheds light on the fact that these systems can't right now.'
According to the researchers, this deficiency could be explained by the way in which AI neural networks are designed. These are mainly inspired by the regions of the human brain that process static images, whereas dynamic social scenes call on other brain areas. This structural discrepancy could explain what the researchers suggest could be 'a blind spot in AI model development.' Indeed, 'real life isn't static. We need AI to understand the story that is unfolding in a scene,' says study coauthor, Kathy Garcia.
Ultimately, this study reveals a profound gap between the way humans and AI models perceive moving social scenes. Despite their computing power and ability to process vast quantities of data, machines are still unable to grasp the subtleties and implicit intentions underlying our social interactions. Despite tremendous advances, artificial intelligence is still a long way from truly understanding exactly what goes on in human interactions. — ETX Studio
Try Our AI Features
Explore what Daily8 AI can do for you:
Comments
No comments yet...
Related Articles
The Sun
2 hours ago
- The Sun
From allies to enemies: the cost of a Musk-Trump split
WASHINGTON: The bitter clash between Elon Musk and Donald Trump has drawn a spotlight on the South African-born tycoon's businesses. Here is a look at his affairs as their White House partnership turns toxic, with billions of dollars in market value and government contracts hanging in the balance. - Tesla - The Tesla car company is the cornerstone of Musk's business empire and has suffered considerably since the entrepreneur dove into politics. The electric vehicle giant's stock has plummeted more than 20 percent since the start of the year, reflecting investor anxiety about Musk's increasingly polarizing public persona. The damage reached a fever pitch on Thursday when the Musk-Trump feud erupted out in the open. In a matter of hours, Tesla shed more than $150 billion in market capitalization, wiping $34 billion from Musk's personal fortune. The alliance with Trump was supposed to have been Tesla's golden ticket, even if the administration was going to scrap tax credits that had helped it become an automobile juggernaut. More importantly, Musk could count on Trump's blessing for his ultimate vision: putting fully autonomous vehicles on American roads. This ambition has been stymied by government regulation over the years, with authorities slowing efforts due to worries that the technology is not ready to hit the road at mass scale. The Trump administration was expected to lift these regulatory constraints -- a promise now in serious jeopardy. 'Musk needs Trump because of the regulatory environment, and you can't have Trump go from friend to foe,' said analyst Dan Ives of Wedbush Securities. The administration also regulates vehicle design and would influence the mass production of robotaxis that Musk intends to launch in a pilot program in Austin, Texas, this month. Musk's hard-right political pivot has alienated the very customers Tesla needs most: environmentally conscious and liberal-leaning buyers who once saw the brand as aligned with their values. Some drivers have resorted to bumper stickers declaring their cars were purchased 'before Elon went crazy.' The damage is showing up in sales figures. In Europe, while overall electric vehicle sales climbed, Tesla's market share crashed 50 percent in April as attention focused on Musk's political activities and the company's aging product lineup. A recent Morgan Stanley survey said 85 percent of investors believe Musk's political involvement is actively harming Tesla's business. - SpaceX - A prolonged battle with Trump poses existential risks for SpaceX, Musk's space exploration company that has become NASA's most critical partner. SpaceX and NASA are deeply interdependent. SpaceX depends on government contracts worth tens of billions of dollars, while NASA relies on SpaceX for everything from astronaut transportation to satellite deployment. SpaceX's portfolio includes some of the most sensitive national security projects: launching astronauts to the International Space Station, building spy satellites and operating the Starlink satellite network. The financial windfall has been enormous, with a December share sale valuing SpaceX at $350 billion -- $140 billion more than just six months earlier, largely due to anticipated government largesse under Trump. In the heat of the clash on Thursday, Trump threatened to cut off all government contracts, while Musk said he would mothball the Dragon spacecraft, which is vital for ferrying astronauts to and from the International Space Station -- though he later walked back this threat. - xAI - Musk has huge plans for his xAI artificial intelligence company. He's angling to compete with OpenAI, the ChatGPT-maker that was co-founded by Musk a decade ago and is now steered by his archrival Sam Altman. Altman has his own inroads to the White House, where he signed a massive AI infrastructure initiative called the Stargate Project, which recently expanded to Saudi Arabia and Abu Dhabi. Initially dismissing Stargate as unrealistic, Musk later worked behind the scenes to undermine the project, reportedly telling investors that Trump wouldn't approve any expansion that excluded xAI. Adding another layer of complexity, Musk folded X (formerly Twitter) into xAI earlier this year. Musk's $44 billion acquisition in 2022 transformed the site into the go-to platform for conservatives, but Trump himself remains an infrequent user, preferring his own Truth Social platform for communication.
New Straits Times
14 hours ago
- New Straits Times
Experts call for relook on local STEM education as US tightens foreign student policies
KUALA LUMPUR: The tightening of foreign student policies in the United States may have disrupted the study plans of many worldwide, but Malaysian experts say this is an opportunity to relook at support and funding for local institutions. Former deputy higher education director-general Professor Datuk Dr Noor Azizi Ismail said with the right ecosystem, greater autonomy and stronger funding, local universities could offer high-quality education in science, technology, engineering and mathematics (STEM) subjects. "We have the talent. Many of our academics are graduates of world-class universities, but they also possess a better understanding of our local context," he said to the New Straits Times. Noor Azizi, who is also former UMK vice-chancellor and Malaysian Graduate School Entrepreneurship and Business lecturer, acknowledged that the US remained home to top STEM institutions like the Massachusetts Institute of Technology (MIT), Harvard and Stanford. But he said the current restrictions could steer students toward equally strong alternatives. "In fact, our local universities may even be more relevant in addressing regional challenges and industry needs," Noor Azizi added. International Islamic University Malaysia (IIUM) principal director (Academic and Internationalisation) Professor Dr Amir Akramin Shafie echoed the view, saying that "brain stay" could strengthen domestic institutions and fill talent gaps, particularly in Malaysia's key sectors, like semiconductors and advanced manufacturing. Amir Akramin, who is also a lecturer at IIUM's Mechatronics Engineering Department, highlighted the strengths of British and Australian universities, with the former known for its excellence in fundamental sciences and engineering, while the latter led in environmental science, mining and biotechnology. "The perception that the US is the sole leader in STEM might be partially outdated," he said. Amir added that Germany, Japan and South Korea had also invested heavily in STEM education and could now offer globally respected programmes. While some students might initially struggle to shift away from the "US dream", Amir Akramin believed many would adapt and develop new preferences. "This will enhance our diversity and increase the overall international experience of our STEM workforce," he said. Both academics also said with the current US policy, the country might experience reduced exchange of ideas and fewer international collaborations, which would eventually shift the global innovation landscape away from American institutions. "On the upside, a shift from "brain drain" to "brain stay" could help Malaysia strengthen local innovation, build institutional capacity and establish itself as a regional hub for STEM education," said Noor Azizi. On May 27, US Secretary of State Marco Rubio ordered a suspension of student visa processing as President Donald Trump's administration ramped up vetting of their social media. It was the latest move after Rubio rescinded hundreds of visas and the Trump administration moved to bar Harvard from admitting non-Americans. Rubio had used an obscure law that allowed the removal of foreigners for activities deemed counter to US foreign policy interests. The most visible targets had been students involved in activism over Gaza. Trump administration officials accused the students of anti-Semitism, charges strenuously denied by a number of those targeted.
The Star
2 days ago
- The Star
China is working on an ultra-fast torpedo powered by AI for submarine warfare
In the recent Chinese blockbuster Operation Leviathan , an American nuclear submarine uses hi-tech acoustic holograms to bamboozle Chinese torpedoes and their human operators. Months after the film hit cinema screens, military researchers in China revealed they were working on an artificial intelligence system designed to cut through exactly this type of underwater deception. In a peer-reviewed paper published in Chinese-language journal Command Control & Simulation in April, the team from the PLA Navy Armament Department and China State Shipbuilding Corporation said their system had unprecedented accuracy for torpedoes travelling at high speeds. Tested against data from classified high-speed torpedo ranges, the technology achieved an average 92.2 per cent success rate in distinguishing real submarines from decoys even during tense exchanges, according to the paper. That is a leap from the legacy systems that often miss the target. Future submarine warfare hinges on deceiving torpedoes using illusions. Hi-tech decoys – as dramatised in Operation Leviathan – are used to replicate a vessel's acoustic signature, generate a false bubble trail to make it look like it is making an emergency turn, or deploy in coordinated swarms to project ghost targets across sonar screens. These tactics are particularly effective against what is known as ultra-fast supercavitating torpedoes – weapons that generate cavitation, or vapour bubbles, around their hulls to reduce drag. The resulting roar drowns out genuine target echoes while distorting acoustic fingerprints, according to the Chinese researchers. 'Current target recognition methods for China's underwater high-speed vehicles prove inadequate in environments saturated with advanced countermeasures, necessitating urgent development of novel approaches for feature extraction and target identification,' said the team led by senior engineers Wu Yajun and Liu Liwen. 'Only those underwater high-speed systems equipped with long-range detection capabilities and high target recognition rates can deliver sufficient operational effectiveness,' they added. The solution they proposed came from an unorthodox combination of physics and machine learning. Facing scarce real-world combat data, the team began by simulating decoy profiles using hydrodynamic models of bubble collapse patterns and turbulence. To do that they used raw data collected from the PLA Navy's high-speed torpedo test range. These simulations were then added to a 'generative adversarial network' – a duelling pair of AI systems. One of them, the generator, refined decoy signatures by studying submarine physics and acoustic principles. Its opponent, the discriminator, trained to detect flaws in these forgeries using seven layers of sonic pattern analysis. After many rounds of training, the system had created a huge collection of artificial decoy profiles. The AI uses a specialised neural network architecture inspired by image recognition, according to the paper. Sonar signals go through a process where they are normalised for amplitude, filtered through correlation receivers to suppress noise, and finally rendered as spectral 'thumbnails' using a mathematical tool known as a Fourier transform. These sonic snapshots then pass through convolutional layers in the neural network that are tuned to detect anomalies in frequency modulation. Pooling operations then average out distortions like bubble interference. The team said that when confronted with the most sophisticated type of decoys, detection rates went from 61.3 per cent to more than 80 per cent. It comes amid a global race to develop 'smart' torpedoes. Russia's VA-111 Shkval torpedo and its US counterparts under development all rely on supercavitation at present, and they struggle with target discrimination at extreme speeds. 'With continuous advancements in modern underwater acoustics, electronic technologies and artificial intelligence, today's underwater battlespace often contains multiple simultaneous threats within a single operational area – including decoys, electro-acoustic countermeasure systems, electronic jammers and diverse weapon systems,' the paper said. In such intense underwater environments where multiple targets or decoys can appear simultaneously, these systems must be able to instantly distinguish authentic targets from false ones to avoid mission failure or a wasted trajectory and to prioritise the highest-threat targets, according to the team. 'Critically, given the autonomous nature of underwater high-speed vehicles, all decisions must be made without real-time external communication support, substantially increasing algorithmic complexity and computational demands,' the team said. 'The deep-learning recognition model proposed in this study, combined with the generative adversarial networks' small-sample identification solution, enables effective underwater target discrimination. This lays the technical groundwork for field deployment,' they added. – South China Morning Post
