explainer
The introduction of ChatGPT’s audio and image capabilities drew mixed reactions online.
OpenAI brings audio and image capabilities to ChatGPT.
The platform has long been limited to written prompts, but it plans to add new features over the next two weeks. Paid version OpenAI announces app development blog post on monday.
Other users will also receive this feature “soon.”
What can I do with the ChatGPT update?
Users can have voice conversations with chatbots, bringing them closer to popular AI assistants like Apple’s Siri and Amazon’s Alexa.
ChatGPT’s new voice capabilities can also narrate bedtime stories, settle arguments at the dinner table, and read text typed by users out loud.
OpenAI says the technology behind it is used by Spotify to translate the platform’s podcasters’ content into different languages.
Users can also upload one or more images to the interface and use drawing tools to highlight specific parts of the image.
Vision capabilities can be used to “troubleshoot why your grill won’t start, inspect the contents of your refrigerator to plan meals, and analyze complex graphs of work-related data.”
How did people react?
OpenAI’s announcement drew mixed reactions on X (formerly Twitter). While some users celebrated the new update, others expressed concern.
While this may sound interesting, I certainly hope that rapid advances in technology and artificial intelligence do not lead to a situation reminiscent of the horrors of Y2K or a potential machine uprising. It is essential for us to develop and manage these responsibly…
— Christopher CSI (@CSI9ja) September 25, 2023
Professor at the University of California, Berkeley Prompt AIhas described the fear of AI becoming too human-like as the “uncanny valley gap.”
While the added features may make chatbots feel more natural, some say a complex interface that cannot mimic human interaction can feel strange and make the product difficult to use. suggested by research.
Users have expressed concern over recent lawsuits by OpenAI for copyright violations and intellectual property infringement, and are advising other users not to use ChatGPT.
ChatGPT is currently under federal investigation for data breaches, inaccuracies, privacy violations, deceptive practices, and reputational harm. Not recommended for use.https://t.co/1EdNKbqFPWhttps://t.co/57rlB3kMFDhttps://t.co/SMF62KRsSehttps://t.co/IBTW92I7FJ
— Nicole Miller (@JOSourcing) September 25, 2023
Some have also mentioned how this update could replace small AI startups, software engineers, and even educators in the future.
So… how many startups have died in the last 5 minutes?
— Terry Tan (@terrytjw) September 25, 2023
@felixchin1 This is essentially what I was describing. The camera is always on and the AI monitors everything and interacts with you like a tutor. If they release this, education as we know it will end.
— Brad (@Brad08414464) September 25, 2023
Software engineers are just the digital plumbers of this field.
— Andy (@AndyTech99) September 25, 2023
AI-generated voices also raise the threat of deepfakes, voice fraud, and identity theft.
of malicious use AI voice generators are on the rise, imitating real human voices and calling relatives to ask for money.a mcafee report It suggests that 77% of people targeted by AI voice scams ended up losing money.
Additionally, adding voice recognition could make the feature less accessible to people who don’t speak in mainstream accents, he said. joel fisherHe studies human-computer interaction at the University of Nottingham, UK.
Since the image feature allows AI to recognize images, users are concerned that bots can bypass image verification CAPTCHA tests on websites.
These tests, which require users to prove they’re not a bot by transcribing distorted text or recognizing images, are designed to limit access.
a recent researchhas not yet been peer-reviewed, but shows that AI bots can solve CAPTCHA tests faster and more accurately than humans.
RIP capture
— Chase (@ChaseMc67) September 25, 2023
Is ChatGPT aware of these risks?
OpenAI is Admitted The voice functionality in the new update could potentially allow malicious attackers to conduct fraud or spoofing. To get around this, the company said it is “using this technology to power specific use cases.”
This is a voice chat with voice actors that the company worked directly with.
The company also acknowledges the limitations of using images with AI, such as hallucinations where AI generates false information about images.
To combat this, OpenAI has taken technical steps to limit ChatGPT’s ability to analyze and speak directly about people.