tech ethics

WTF Fun Fact 13625 – AI and Realistic Faces

Researchers at The Australian National University (ANU) have found that AI-generated faces now appear to be more realistic faces than those of actual humans. But that’s only true if the AI is generating the faces of white people.

This development raises crucial questions about AI’s influence on our perception of identity.

Training Bias in AI

This study reveals a concerning trend. People often see AI-generated white faces as more human than real ones. Yet, this isn’t the case for faces of people of color.

Dr. Amy Dawel attributes this to AI’s training bias. AI algorithms have been fed more white faces than any other. This imbalance could increase racial biases online. It’s especially troubling in professional settings, like headshot creation. AI often alters skin and eye colors of people of color, aligning them more with white features.

The Illusion of AI Realistic Faces

Elizabeth Miller, co-author of the study, highlights a critical issue. People don’t realize they’re being fooled by AI faces. This unawareness is alarming. Those who mistake AI faces for real ones are often the most confident in their judgment.

Although physical differences between AI and human faces exist, they’re often misinterpreted. People see AI’s proportionate features as human-like. Yet, AI technology is evolving rapidly. Soon, distinguishing AI from human faces could become even more challenging.

This trend could significantly impact misinformation spread and identity theft. Dr. Dawel calls for more transparency around AI.

Keeping AI open to researchers and the public is essential. It helps identify potential problems early. Public education about AI’s realism is also crucial. An informed public can be more skeptical about online images.

Public Awareness and Tools for Detection

As AI blurs the line between real and synthetic, new challenges emerge. We need tools to identify AI imposters accurately. Dr. Dawel suggests educating people about AI’s realism. Such knowledge could foster skepticism about online images. This approach might reduce risks associated with advanced AI.

ANU’s study marks a significant moment in AI development. AI’s ability to create faces now surpasses human perception in certain cases. The implications are vast, touching on identity and the potential for misuse.

As AI evolves, transparency, education, and technological solutions will be key. We must navigate these challenges collectively to ensure AI’s responsible and beneficial use.

— WTF fun facts

Source: “AI faces look more real than actual human face” — ScienceDaily

WTF Fun Fact 13613 – First Chimeric Monkey

Researchers have made a monumental stride in primate research by making the first chimeric monkey.

This marks the first successful birth of a chimeric monkey from embryonic stem cell lines. This scientific achievement has profound implications for the fields of genetic engineering, species conservation, and biomedical studies.

Understanding Chimerism in Primates

The study, led by senior author Zhen Liu of the Chinese Academy of Sciences, culminated in the birth of a monkey with cells originating from two distinct embryos. Until now, this feat of chimerism had been achieved only in smaller mammals such as rats and mice. Published in the prestigious journal Cell, the research opens new avenues for understanding pluripotency. That’s the capability of stem cells to differentiate into any cell type—in non-human primates and possibly humans.

The cynomolgus monkeys, commonly used in biomedical research, served as the subjects for this groundbreaking experiment. The researchers established nine stem cell lines from blastocyst embryos and selected a subset of these pluripotent cells to inject into early-stage monkey embryos. This meticulous process led to several pregnancies and the birth of six live monkeys. One of these showcased a substantial level of chimerism.

The Making of a Chimeric Monkey

The researchers tagged the stem cells with green fluorescent protein. This enabled them to trace which tissues originated from the stem cells. Extensive analysis revealed that the chimeric monkey exhibited a wide distribution of stem-cell-derived tissues across the brain, heart, kidney, liver, and gastrointestinal tract. Remarkably, the live monkey displayed stem cell contributions ranging from 21% to 92% across various tissues, averaging 67%.

The presence of stem-cell-derived cells in the reproductive tissues was a significant discovery. It underscors the potential for these cells to contribute to the germline and possibly influence future generations.

Implications and Future Directions

The success of this study is not merely academic. It has practical implications, offering the potential to create more precise monkey models for neurological and other biomedical research. By enhancing the understanding of primate cell developmental potential, the study paves the way for innovative approaches in medical science.

Looking ahead, the team aims to refine their method to increase the efficiency of generating chimeric monkeys. They plan to optimize the stem cell cultures and the blastocysts’ environments, hoping to improve the survival rates of these embryos in host animals.

In conclusion, the birth of the first chimeric monkey from embryonic stem cells is a remarkable scientific milestone. It broadens our knowledge of primate biology and holds promise for future applications that could benefit both primate conservation and human health.

— WTF fun facts

Source: “First live birth of a chimeric monkey using embryonic stem cell lines” — ScienceDaily

WTF Fun Fact 13431 – Ford’s Self-Repossessing Truck

In a move that could revolutionize the automotive industry, Ford has patented a system enabling self-repossessing trucks. Through connectivity and automation, these vehicles could potentially drive themselves back to the dealer in the event of missed payments.

This idea of autonomous repossession is a powerful example of how technology is reshaping our world.

Self-repossessing trucks in the patent stage

The patent, as reported by The Verge, illustrates a system in which a vehicle can determine if it’s subject to repossession based on data received from a remote source. It’s not yet a reality.

However, in the future, a car could disable its own engine or even navigate itself back to the lender, provided it’s safe and legal to do so.

While this might sound dystopian to some, it’s essential to consider the problem this solution aims to address. Traditional repossession methods are far from perfect. They can lead to conflicts, damages, or, in some cases, even legal issues. Autonomous repossession could potentially eliminate these complications, making the process smoother for all parties involved.

But wait, there’s more!

But, Ford’s self-repossessing trucks patent doesn’t stop there. The documentation suggests that these vehicles could also potentially find and connect with a tow truck for assistance if self-driving to the dealer isn’t possible. The vehicle’s systems could even schedule a convenient pick-up time with the towing company, streamlining the entire process.

However, such a concept isn’t without its concerns. Critics argue that autonomous repossession could pose privacy risks. After all, the technology requires tracking and controlling vehicles remotely, raising questions about data security and personal privacy. Ford, along with all automakers pursuing such technologies, will need to address these concerns as development progresses.

Will it really work?

Furthermore, the actual implementation of this system in real-world conditions is another challenge. Various laws regulate vehicle repossession, which differ from state to state, and the legality of a self-repossessing car is still untested. Therefore, while the patent is granted, the road to seeing self-repossessing cars in action might be long and winding.

Ford’s self-repossessing car patent is a clear reflection of our changing times. It encapsulates both the remarkable potential of automation and connectivity, and the complex challenges that come along. It’s a compelling narrative of how technology continues to reimagine our world and how we must navigate the balance between innovation and privacy, efficiency and legality.

— WTF fun facts

Source: “Ford’s self-repossessing car patent is a nightmare of the connected-car future” — The Verge

WTF Fun Fact 13353 – New Deepfake Research

New deepfake research is attempting to resurrect victims of crime for the common good. Of course, like all deepfakes, this too raises ethical concerns.

How new deepfake research brings people “back from the dead”

Deepfakes are a technology that uses artificial intelligence to create hyper-realistic images and videos of people. But so far, they’ve largely been misused to spread misinformation.

Now, researchers at the University of Florida and the University of Michigan are now exploring the possibility of using deepfake resurrections to promote the public good.

Their study focuses on “deepfake resurrections.” This refers to bringing deceased individuals virtually “back to life” using AI-generated images and videos. The researchers emphasize that their approach is different from controversial cases of deepfake resurrections, such as the ones used for political manipulation or commercial purposes. Instead, they aim to explore scenarios where deepfake resurrections could have a positive impact on society.

Can this technology be used for good?

The researchers conducted a study involving approximately 2,000 participants to explore the potential applications of deepfake resurrections. In this study, they focused on creating deepfake resurrections of victims of drunk driving and domestic violence. The aim was to examine the reactions of the participants and assess whether such resurrections could effectively raise awareness about these pressing social issues.

By using deepfake resurrections to share the stories of these victims, the researchers sought to humanize the issues and evoke empathy in the audience.

However, the PSA had little effect and a more negative than positive reaction. The researchers chalked this up to the lack of trust in deepfakes overall, noting that this affected the effectiveness of deepfake resurrections in raising awareness about social issues.

Ethical considerations

The exploration of deepfake resurrections for the public good raises several ethical questions. One major concern is consent. Since the deceased cannot provide consent, the researchers suggest obtaining permission from the deceased’s estate or family members. This would require creating guidelines to ensure that deepfake resurrections are used in a manner that respects the individual’s legacy and values.

Another ethical consideration is the potential emotional impact on the deceased’s loved ones. While some may find comfort in deepfake resurrections, others might perceive it as a disturbing or disrespectful act. To address this concern, researchers propose involving mental health professionals in the development of deepfake resurrections to ensure they are created with sensitivity and empathy.

Lastly, there is the question of authenticity. The researchers acknowledge the potential for deepfake resurrections to spread misinformation or perpetuate false narratives. To mitigate this risk, they suggest that deepfake resurrections should be transparently labeled as such and accompanied by disclaimers.

— WTF fun facts

Source: “Dying To Tell You: “Deepfake Resurrections” To Promote Public Good Explored By Researchers” — IFL Science

WTF Fun Fact 13333 – Dead Bird Drones

In a somewhat morbid turn of events, scientists are reportedly making dead bird drones. Researchers at the University of New Mexico are using taxidermy techniques to preserve dead birds in order to outfit them with drone components, such as motors, propellers, and sensors.

Why dead bird drones?

Drones have long been modeled on birds and their flight mechanics. In this sense, it’s no surprise that engineers would try to improve drone design using birds so their machines can navigate through difficult terrain and adapt to changing conditions with ease. But the zombie bird thing gives us pause.

However, the project is still in its early stages. the researchers are experimenting with different types of birds to see which ones make the best drones. So far, they have tested the concept using quails and doves.

They’ve found that bird drones can fly and maneuver in ways that traditional drones cannot. For example, bird drones are able to fly close to the ground and through narrow gaps. This could eventually be useful for tasks such as search and rescue or inspecting infrastructure.

Isn’t this morbid?

While the concept of turning dead birds into drones may seem macabre, the researchers behind the project believe that it could have important implications for the future of drone technology. For example, bird-like drones are more energy-efficient than traditional drones.

“If we learn how these birds manage … energy between themselves, we can apply (that) into the future aviation industry to save more energy and save more fuel,” said Dr. Mostafa Hassanalian, a mechanical engineering professor who is leading the project at the New Mexico Institute of Mining and Technology in Socorro.

Bird drones can also be less obtrusive than traditional drones, which may be noisy and disruptive. Since they’re quieter, bird drones could even be used for wildlife monitoring or scientific research without disturbing the animals being studied.

Of course, there are still many challenges to overcome, including the ethical considerations of taking an animal (even a dead one) and turning it into a tool.

Right now, the taxidermied prototype the engineers have built can only fly for 20 minutes.

— WTF fun facts

Source: “Scientists Turned Dead Birds Into Zombie Drones to Spy on Humans” — Popular Mechanics

WTF Fun Fact 13231 – The Downsides of ChatGPT

ChatGPT may seem like the answer students, coders, and writers have been looking for, but it’s far from perfect. In fact, the generative AI can be riddled with misinformation that’s likely to get you caught. And the downsides of ChatGPT are easy to discern since you can just ask the chatbot itself.

What is ChatGPT?

Chat GPT is a chatbot built by Open AI. Just go to chat.openai.com and see for yourself. You can ask it anything.

Of course, that doesn’t mean you’ll get the right answer. Sure, it may be the answer you want to hear. It may be grammatically correct. It may even seem elegantly stated. But look a little closer, and you’ll find it’s relatively simplistic and repetitive. And it can be downright incorrect and misleading.

What are the downsides of ChatGPT?

We went right to the source to find out how the AI might come up short. According to ChatGPT itself (when asked “what is one of the major drawbacks of chatgpt?”):

“One of the major drawbacks of ChatGPT is its lack of context awareness and the possibility of generating biased or incorrect responses.

Since ChatGPT is trained on a large corpus of text data from the internet, it may sometimes generate responses that reflect the biases and inaccuracies present in that data. This can lead to the generation of incorrect information or responses that perpetuate harmful stereotypes or beliefs.

Additionally, ChatGPT operates on a token-by-token basis, meaning that it doesn’t have an overall understanding of the context of a conversation. This can lead to responses that are not fully relevant or coherent, especially when the input is complex or requires a deep understanding of a particular subject.

To mitigate these issues, it is important to carefully evaluate and curate the training data used to train language models like ChatGPT, and to use appropriate post-processing techniques to ensure the quality and accuracy of the generated responses.”

We’re impressed the AI went the extra mile and suggested some more shortcomings!

The future of ChatGPT

Yes, the chatbot is a viral sensation. And it absolutely has the ability to do some good and save some time. But it also has the ability to be used to fill the Internet (and your homework assignments) will all sorts of confident-sounding information that’s factually incorrect. And the kicker is that it may someday be trained on the same incorrect websites it helped create.

Of course, there are some pretty ominous things ChatGPT can be used for. And to be fair, Open AI has been…well, open about the downsides of ChatGPT. According to Bleeping Computer (cited below), “OpenAI has been upfront about the AI’s shortcomings including its ability to ‘produce harmful instructions or biased content and continues to fine-tune ChatGPT.” — WTF fun facts

Source: “OpenAI’s new ChatGPT bot: 10 dangerous things it’s capable of” — Bleeping Computer