- The Daily Update
- Posts
- New deepfake protection tech
New deepfake protection tech
Study: Gen AI benefits low-performing workers the most
Jack Shields
November 13, 2023
Welcome back!
Celebrities and AI experts are concerned that social media platforms aren’t ready for the rise of AI deepfakes. Now, researchers are developing a new tool that may protect people from deepfake abuses.
Let’s jump right into today’s top stories.
In today’s Daily Update:
🗞️ New tech helps protect people from deepfakes
🤖 Generative AI benefits low-performing workers the most
📸 Google in talks to invest in Character.AI
🚨 AI Roundup: Three quick hits
Read time: 2 minutes
LATEST NEWS
🗞️ New tech helps protect people from deepfakes
Source: Adobe Firefly
Researchers at Washington University in St. Louis are developing a new tool called AntiFake that can help protect people from deepfake abuses.
What you need to know:
The team led by Ning Zhang will present the project at a major security conference in Denmark later this month.
AntiFake scrambles the audio signal of a voice track to confuse AI models. The modified track sounds normal to the human ear, but makes it more difficult for AI to create voice clones.
The tool won’t protect people whose voices are already widely available online.
AntiFake was inspired by Glaze — a similar tool that protects visual artists from having their works scraped for AI models.
The relevance: AI-generated synthetic media is already starting to flood the internet and social media. Several celebrities have had their likenesses stolen by AI clones to promote products without their consent. AntiFake offers a promising solution for ordinary people who want to protect themselves from AI deepfakes.
RESEARCH SPOTLIGHT
🤖 Generative AI benefits low-performing workers the most
A new study published by the Boston Consulting Group shows AI tools like ChatGPT boost low-performing employees the most.
The details:
The study examined AI’s effects on 758 of Boston Consulting Group’s management consultants.
Prior to the exercise, participants were divided into “top-half” and “bottom-half” skill groups. All participants then completed a financial consulting project.
Bottom-half participants who were assisted by ChatGPT improved their performance by an average of 43%.
Top-half participants improved their performance by 17% with ChatGPT.
Why it matters: The previous consensus among researchers appeared to be that AI would aid top performers and push lower-skiled workers out of jobs. This research indicates that generative AI tools could actually be an equalizer for less productive workers.
BUSINESS SPOTLIGHT
📸 Google in talks to invest in Character.AI
Screenshot: Character.AI
Google is reportedly in talks to invest hundreds of millions of dollars in Character.AI. This would deepen an existing partnership between the two companies.
Key points:
Character.AI allows users to create personalized chatbots and AI assistants.
Some of the platform’s most popular characters are virtual versions of celebrities like Elon Musk, Mark Zuckerberg and Kanye West.
Character.AI has attracted 100 million monthly website visits in the first six months since its launch. Users aged 18 to 24 make up about 60% of this traffic.
The significance: The Tech Giants are heavily investing in AI chatbots with humanlike personalities. Character.AI’s chatbots seem similar to Meta’s AI personas that launched in late September. The trend toward AI chatbots designed for natural conversations and social interactions is intriguing to say the least.
MORE TRENDING NEWS
🚨 AI Roundup: Three quick hits
Source: Adobe Firefly
Level Up: OpenAI CEO Sam Altman says the upgraded GPT-4 Turbo model is officially live in ChatGPT.
AI Video Upgrade: Runway debuts new “Motion Brush” tool that allows for improved movement control.
Fighting Back: Google files lawsuit against scammers who are encouraging users to download an “unpublished” version of Bard that’s actually malware.
THAT’S ALL FOR TODAY