Hold the Code’s 48th edition features dives into machine ownership of art and AI involvement in gun policy. Our weekly feature highlights a new Intelligent Ultrasound that reduces much of the risk for some types of surgery. Be sure to check out HTCs crossword scavenger hunt and win free Tomate! In case you need a little hint, our Editor Dwayne’s favorite food is Chicken! As always, happy reading and be sure to share HTC with a friend!
Dalí or DALL-E: Who owns AI art?
Source: Craiyon
From generating memes to producing the June cover of Cosmopolitan, AI has accomplished some impressive artistic feats.
Today, popular AI art-generation tools include DALL·E, initially released by OpenAI in 2021, and Craiyon, released by open-source AI-company Hugging Face. Together, these platforms receive over 1.5 million users per day, generating thousands of images every minute. The creation process is simple: users input text describing the image they want, and the AI tools will use a database of billions of pictures found on the Internet to generate new images
While this new surge in AI’s artistic capabilities opens the door to exciting possibilities in content creation, it also raises concern about the complications it might bring. After all, when a machine is making art, who should be responsible for it?
Inspiration or imitation?
AI creates new images by drawing directly from existing images, many of which were created by human artists. As a result, some artists are worried about the potential for widespread use of their work without appropriate credit.
Greg Rutkowski, an artist from Poland, has had over 90,000 AI images created using his name, resulting in a plethora of digital images emulating his distinctive fantastical art style.
“It’s been just a month. What about in a year? I probably won’t be able to find my work out there because [the internet] will be flooded with AI art. That’s concerning” said Rutkowski.
Can AI own art?
However, attributing ownership for AI-generated images can be a complicated matter. Does the artwork belong to the AI tool, the people who used the tool, the people who created the original images, or some other party altogether?
On one hand, there is little legal precedent for dealing with AI-generated works, as many countries only have copyright laws that apply to humans. On the other hand, companies like OpenAI have announced policies that would allow for commercial use of images created with DALL·E. Already, a report by PetaPixel found that AI-generated images are being sold on major stock photo platforms like Adobe Stock.
As AI’s impact on art grows, the need for clear standards on image rights becomes more urgent. “It’s a cool experiment,” Rutkowski noted. “But for me and many other artists, it’s starting to look like a threat to our careers.”
Too good to be true: Gunshot tech draws controversy
Source: Wikimedia Commons. Image via Noah Wulf
Politicians want a quick fix to the issue of gun violence in big cities. For a while, they thought they had it.
ShotSpotter is an artificial-intelligence-powered gunshot detection tool that has been implemented in many big U.S. cities, with their two biggest customers being New York City and Chicago. Other cities, including Detroit, Seattle and Portland, are either considering or are in the process of implementing technology like ShotSpotter as they scramble to address an increase in violence in their cities.
Let’s slow down for a second. How does this tech work?
How does ShotSpotter work?
When a city signs a contract with ShotSpotter, the following occurs:
ShotSpotter places sensors that identify loud noises
Loud noise sent to ShotSpotter for analysis
If the noise is identified as gunfire, ShotSpotter alerts city police department
Officers dispatched to the scene
The purpose of gunshot detection technology is to speed up police response, especially in situations where no one has called 911.
The company says that its technology is not meant to reduce homicides or violence overall. Instead, it can help law enforcement find victims and collect evidence faster.
“Unless the state allows cities to set our own gun laws, this technology is one of few new options available to enhance current efforts,” said a spokesperson for the Seattle mayor
However, the technology is not without controversy.
The Controversy
Chicago is an integral source of the ShotSpotter debate, as the city has been using ShotSpotter since 2017. A 2021 study by the MacArthur Justice Center in Chicago found that the police officers responding to a ShotSpotter alert did not find a gun-related crime 90% of the time.
While the company maintains that its technology is accurate 97% of the time, these stats suggest otherwise.
The MacArthur Justice Center is leading a pending class action lawsuit against the city of Chicago. The residents they work with say that they were wrongfully arrested by Chicago law enforcement as a result of ShotSpotter technology. Jonathan Manes, an attorney with the MacArthur Justice Center, says the gunshot detection technology can cause more harm than good.
“Everybody wishes that there was a technological silver bullet to address the problem of gun violence… when in fact there’s very little evidence that this has helped,” said Jonathan Manes
Weekly Feature: Under Your Nerves: AI Medical Vision
Source: Intelligent Ultrasound
Undergoing surgery is often both physically straining and mentally taxing for patients. Advances in pain-relief technology have made major medical procedures possible, yet inefficiencies in the process of anesthesia administration continue to show their effects.
General anesthesia can simplify operations, yet they can be a risky choice for those with respiratory-related conditions, and they have a string of potential postoperative side effects. Intelligent Ultrasound’s newly introduced ScanNav AI technology aims to improve the accuracy of Peripheral Nerve Block (PNB) injections, which act as an alternative to general anesthesia and central nerve blocks.
What are PNBs?
Peripheral Nerve Block (PNB) injections are a form of regional anesthesia in which an anesthetic is injected near the nerves that are associated with the area to be operated upon. When compared to General Anesthesia:
PNBs can often lessen postoperative pain/fatigue
Reduce the amount of opiate/pain medication needed after surgery
Allow patients to be awake during an operation
Compared to other forms of local/regional anesthesia, PNBs can also be used for smaller or more precise nerve targets.
Nonetheless, PNBs have their own disadvantages: they require much more precision when administering the injection, which often necessitates the usage of external equipment such as ultrasound to locate the optimal injection site. This process is difficult, and can result in nerve block failure and trauma to the surrounding anatomical structures if done incorrectly.
What is ScanNav?
ScanNav is a real-time AI software that can analyze an ultrasound image and highlight nerves and important anatomical structures prior to injection. ScanNav also creates 3D visualizations of the area, allowing doctors to more accurately identify the optimal path to the peripheral nerve being targeted for injection.
In testing, the data regarding ScanNav’s performance is incredibly promising:
The AI was able to identify key anatomical structures “in nearly 94% of cases.”
False-positive and false-negative rates “clocked in around 3%.”
ScanNav’s automated overlay helped reduce the risk of nerve block failure by 81%.
The study also showed an 86% decrease in trauma to the surrounding anatomy in the ultrasounds studied.
The AI allows doctors to see the original, unmodified ultrasound image alongside a real-time, AI-generated image with “annotations” that help doctors in their decisions on injection location.
Future Applications
ScanNav was recently granted de novo approval from the FDA for usage of its AI-based technology. It had previously been granted approval in the U.K. and Europe in the spring of 2021, and now plans to bring the product to U.S. markets. Intelligent Ultrasound previously had successfully created its ScanNav Assist tool, which used AI to analyze ultrasound fetal anomaly scans. In collaboration with GE Healthcare, Intelligent Ultrasound is working on integrating its AI directly into ultrasound scanners, potentially making medical procedures that require ultrasound more precise and useful.
Love HTC?❤️
Follow RAISO (our parent org) on social media for more updates, discussions, and events!
Instagram: @Raisogram
Twitter: @raisotweets
RAISO Website: https://www.raiso.org