Discover more from Yaro on Ai and Tech Trends
🌍 Global Trust in AI is Dropping
Plus: The Scary Rise of AI Voice Scams and Instant Fake Document Creation.
by Contentful - The future of intelligent composable content
It’s Friday yeah, Thanks for joining us for the last edition of the week where we share about the scary Rise of AI Voice Scams and Instant Fake Document Creation as well as how AI global trust is dropping. Hope you enjoy it and feel free to ask any questions you may have.
This is what we got today…
Global Trust in AI is Dropping
The Future of Intelligent Composable Content (Contentful)*
The Scary Rise of AI Voice Scams and Instant Fake Document Creation.
Is Claude 3 A better AI Model?
Our Sections:
❓Quiz of the Day
📰 News and Trends
🤔 What I Found Interesting
🧰 AI Tools (Speech II)
❓Quiz of the Day
Which AI milestone involved defeating the world champion in the game of Go?
A) IBM Watson on Jeopardy!
B) Deep Blue vs. Garry Kasparov
C) AlphaGo vs. Lee Sedol
D) OpenAI Five playing Dota 2
The correct answer will be given in the next newsletter - Last Week’s Quiz correct answer: “A”
📰 AI News and Trends
Inflection AI's friendly chatbot tops 1 million daily users (Axios)
Key OpenAI Executive Played a Pivotal Role in Sam Altman’s Ouster (NYT)
D-Wave says its quantum computers can solve otherwise impossible tasks (NS)
Accenture to acquire Udacity to build a learning platform focused on AI (TC)
Google engineer indicted over allegedly stealing AI trade secrets for China (TheVerge)
🌐 Other Tech news
Breaking: Bitcoin Reaches New All-Time Highs to $70,000! (ITrust)
How Hackers Dox Doctors to Order Mountains of Oxy and Adderall
US lawmakers vote 50-0 for the sale of TikTok despite angry calls from users (ARS)
Engineer plans solar panel implant for human retina to retain eyesight (IE)
Temu, Nearly Unheard of a Year Ago, Is Redefining E-Commerce in the U.S. (WSJ)
Global Trust in AI is Dropping
The poll by consultancy firm Edelman, involving 32,000 global respondents, reveals a decline in public trust in AI since 2019. Trust in AI has dropped globally from 61% in 2019 to 53% in the current survey. In the US, where job insecurity concerns related to AI are prevalent, only 35% of people now trust AI, down from 50% five years ago. The majority of people (two-to-one margin) believe that AI innovation has been "badly managed" so far. While 76% trust the tech industry in general, only half trust AI specifically. The public looks to scientists to inform them about AI safety, presenting an opportunity for the research community to establish authority on the subject.
The graph shows that global trust in AI has decreased from 61% in 2019 to 53% in the current survey, while in the US, trust has dropped more significantly from 50% in 2019 to 35% currently.
The Future of Intelligent Composable Content
Contentful brings intelligent digital experiences to life. With generative AI and composable content, the possibilities are as limitless as your team’s creativity.
The Scary Rise of AI Voice Scams and Instant Fake Document Creation
AI is increasingly becoming a tool for criminals to enhance and accelerate their illicit activities, raising concerns at a rapid and alarming pace. The technology now powers sophisticated phone scams, where AI can mimic the voices of your friends and family to deceitfully request money. Additionally, AI facilitates the instantaneous creation of highly realistic fake documents, further enabling fraud and identity theft. This advancement in criminal techniques underscores the urgent need for heightened awareness and stronger protective measures against the misuse of AI technology.
OnlyFake, an underground website, is exploiting neural networks to create highly realistic fake IDs for just $15, posing a significant cybersecurity threat and potentially facilitating activities like bank fraud and money laundering. This service, as verified by 404 Media, enables the creation of fake IDs (like the one below) that are convincing enough to pass some online verification systems, such as those used by cryptocurrency exchanges. OnlyFake's approach marks a departure from traditional methods of producing fake IDs, offering a quicker, less skill-intensive alternative that could produce up to 20,000 documents a day. The simplicity and accessibility of this technology raise concerns about its potential to undermine identity verification processes across various platforms.
Deeper Dive: Implications and Protections
How Dangerous Is This?
The emergence of OnlyFake introduces a substantial risk to both individuals and institutions by making the production of convincing fake IDs accessible to a broader audience. This can lead to an increase in:
- Identity Theft: As their personal information could be used without their consent.
- Financial Fraud: Easier access to fake IDs could lead to an uptick in fraudulent financial activities, including unauthorized account openings and transactions.
- Security Breaches: Institutions may face challenges in authenticating identities, leading to potential breaches in secure environments.
Protecting Against Fake ID Risks:
1. Enhanced Verification: Adopt biometric and advanced verification methods. However, tech companies will have access to sensitive biometrics and personal data.
2. Public Awareness: Educate on protecting personal information.
3. Regulatory Efforts: Increase crackdowns on such sites and tighten verification laws.
4. Individual Caution: Be vigilant for identity theft signs and safeguard personal data.
Addressing this threat requires combined efforts from technology, education, and regulation to ensure security and privacy in the face of advancing fraudulent practices.
📝 Help us help you.
Please help us to bring you better content and a better newsletter by filling out this quick form. Everyone who fills this out will get a FREE 3-month subscription.
Your support helps us help you. Thank you!
Is Claude 3 A better AI Model?
Claude 3, Anthropic's new AI model on par with GPT-4, feels remarkably warm and human-like, making it the most naturalistic AI assistant yet. As a writer, I've found Claude 3 invaluable for creative assistance and editorial suggestions. Its graph generation capabilities are also impressive.
While the free version's chat navigation isn't as smooth as ChatGPT's, prompting Claude 3 is very intuitive overall. In terms of performance, Claude 3 is similar to GPT-4, though it handles larger files better and fosters more natural conversations. Graph creation is also more streamlined.
A key limitation of the free Claude 3 is its inability to generate or manipulate images, though it can analyze existing ones.
Overall, Claude 3 excels as a warm, creative AI companion, with powerful language understanding and data visualization abilities. As AI models advance, Claude 3 represents a significant achievement in naturalistic AI interaction.
Pros of Claude 3:
Feels extremely warm, human-like, and naturalistic in interactions.
Excels as a creative writing companion and editorial aid.
Long context window of 1 million tokens with high accuracy.
May use synthetic data, reducing data constraints.
Lower pricing than comparable models.
Cons of Claude 3:
Doesn't outperform GPT-4 on benchmarks.
Uncertain if Anthropic can deliver a much more powerful model.
Technical advantages like context window may be replicated by competitors.
Reliance on cloud partners like AWS and Google for distribution.
📰 Publications I am currently reading and recommending:
1-Minute Question - Based on Neuroscience Science Practices to rewire your brain for success.
How Brands Win - Distills the strategies behind great brands. And gives you actionable insights for your business.
Paper Planes - Single men and women who are growth-minded, willing to work on themselves, and open to spiritual concepts like manifestation
🧰 AI Tools
Speech II
podcast.ai - A podcast that is entirely generated by artificial intelligence, powered by Play.ht text-to-voice AI.
VALL-E X - A cross-lingual neural codec language model for cross-lingual speech synthesis.
TorToiSe - A multi-voice text-to-speech system trained with an emphasis on quality. Opensource
Bark - A transformer-based text-to-audio model. Opensource
Download over 400+ Tools free here.
Subscribe to Yaro on Ai and Tech Trends
Stay on top of the AI and tech trends in 4 Minutes. Embrace the future of tech and digital economies. Join our vibrant community now!