Welcome, AI Enthusiasts!

Today’s AI Insights:

• 👀 Trump's impact on AI regulation

• 🤔 AI's reasoning skills under fire

• 🤖 Turn your PDF into an AI chatbot

• 🗞️ Latest AI and tech news

• 🛠 New AI tools

• 📝 AI Meme of the day

^{Read Time: 3 mintue}

AI NEWS

AI REGULATION

👀 Trump's impact on AI regulation

Trump’s recent election victory could bring changes to AI regulation in the U.S., with a shift away from Biden’s policies and a potentially lighter regulatory approach.

The Details:

• Biden’s AI Executive Order (EO): Biden’s 2023 AI EO set voluntary guidelines for the AI industry, including mandatory reporting on model security, addressing biases, and guidelines on the safe use of AI. It also established the U.S. AI Safety Institute (AISI) under the Commerce Department.

• Trump’s Stance: Trump has expressed plans to dismantle Biden’s AI EO, with allies arguing that Biden’s policies discourage innovation. Critics say the current reporting rules could deter startups and add red tape to tech growth.

• Possible Future Policies: Trump’s past policies favored AI advancement without strict regulation, focusing on free speech and minimal restrictions. His administration may instead focus on AI’s physical risks, like bioweapon prevention, rather than strict usage guidelines.

• State-Level AI Regulations: If federal oversight lessens, state governments might introduce their own regulations, as seen in California and Colorado, which have already implemented AI safety laws.

With AI transforming industries globally, any shift in U.S. policy will impact innovation, safety, and ethical standards across the sector. If the federal approach becomes less stringent, state-level actions may fill the gap, shaping AI’s development in the U.S.

APPLE

🤔 AI's reasoning skills under fire

A new study from Apple’s AI research team has found serious weaknesses in large language models (LLMs) like those from OpenAI and Meta, especially in tasks that require mathematical reasoning.

The Details:

• The study showed that small changes in the way questions were asked led to inconsistent results from AI models, highlighting their difficulty with logical consistency.

• Researchers developed a new benchmark, GSM-Symbolic, to avoid existing data contamination problems and provide more reliable ways to measure the reasoning ability of these models.

• All 20 models tested, including OpenAI’s GPT-4o and Meta’s Llama 3, performed worse when small variables like numerical values or irrelevant details in questions were altered.

• The study found no signs of formal reasoning in these models. Even changing a name in a question could lead to a different and incorrect answer.

Apple’s study highlights the challenges AI models face with logical and mathematical tasks. It suggests that while AI has advanced in many areas, more development is crucial to improve their reasoning capabilities, especially for real-world applications that require consistent and reliable decision-making.

AI & TECH NEWS

NEW AI TOOLS:

🎥 Opus Clip- Turn long videos into viral clips with Opus Clip

🤖 Flot AI- AI Copilot for all apps and websites.

💼 ReachifyMe- Build and monetize your personal brand on LinkedIn with AI.

🖥️ SlidesGPT- AI-powered presentation creation in seconds

💻 CodeDesign- AI website builder that allows users to create stunning and responsive websites with ease

AI TUTORIAL

AI TUTORIAL

🤖 Turn your PDF into an AI chatbot with Chatpdf

ChatPDF is an AI-powered tool that allows you to interact with PDF documents by asking questions and extracting information.

Steps to follow:

1. Go to the ChatPDF website.

2. Click “Drop PDF here” to upload your file or input the PDF’s URL.

3. Wait for the AI to process the document, which might take a few moments depending on the file size.

4. Use the chat interface to ask questions about the document or use suggested queries.

5. The AI will provide responses based on the content of your PDF, helping you quickly gather needed information.

AI & TECH NEWS

Google.org pledges $15M for AI training for government workers

Google has announced $15 million in grants to train U.S. public sector workers in AI skills. Of this, $10 million will go to the Partnership for Public Service to launch a Center for Federal AI in 2025, aimed at training federal employees on responsible AI use. InnovateUS will receive $5 million to expand AI training to state and local government workers, with a goal of reaching 100,000 individuals across 30 states. This initiative addresses skills gaps and aims to improve government services through AI.

Perplexity AI introduces new features for paid customers

Perplexity AI has introduced two new features for its paid users: Internal Knowledge Search and Spaces. Internal Knowledge Search allows users to search both the web and uploaded files simultaneously, enhancing productivity for enterprises. Spaces, a collaboration tool, enables teams to share files and customize the AI Assistant for specific tasks.

Adobe’s project Super Sonic

Adobe’s Project Super Sonic uses AI to create sound effects from text prompts, identify objects in videos to generate relevant audio, and imitate voices to produce background sounds. It aims to streamline audio production by converting simple text instructions into high-quality sound effects, enhancing the video editing process.

Lenovo unveils AI innovations at Tech World 2024

At Lenovo’s Tech World 2024, the company unveiled several AI-driven innovations, including Lenovo AI Now, a local AI agent that transforms PCs into personalized assistants while ensuring strong data protection. Another key release was the ThinkPad X1 2-in-1 Gen 10 Aura Edition, designed for hybrid work, offering enhanced productivity and adaptability for professionals and student.

AI MEME OF THE DAY

SPONSOR US

AI Insights newsletter is read by thousands of AI and Tech professionals/enthusiasts around the world.

Get in touch to get your product seen today!

Or email us at: [email protected]

THANK YOU FOR READING

FEEDBACK

We'd love to hear your feedback or any interesting thoughts you have!

Please share by replying to this email.