Welcome, AI RESEARCHERS.

Google's experimental Gemini 1.5 Pro took the lead in the AI chatbot race today, surpassing GPT-4 and Claude 3.5 by an impressive 14 points.

Is Google the new AI leader, or should we expect a strong counter-response from OpenAI and Anthropic? Let's dive in...

In today’s AI Researchs:

• Google’s Gemini AI Model Tops AI Benchmark Charts

• Major AI Developments this week

• FLUX.1: A New AI Image Generation Contender

• Microsoft and OpenAI: New Frenemies in AI

• Use Runway's Gen-3 Alpha Model

• AI Researchs & AI pointers

• Trending Ai tools

• More Ai news and developments

Read time: 5 minutes

GEMINI AI ADVANCEMENT

🏆 Google’s Gemini AI Model Tops AI Benchmark Charts

AI Researchs: Google's experimental Gemini 1.5 Pro AI model has outperformed OpenAI's ChatGPT-4o in recent benchmark tests, sparking excitement in the AI community.

The Details:

• Gemini 1.5 Pro achieved a score of 1,300 in the LMSYS Chatbot Arena benchmark, surpassing GPT-4o’s 1,286 and Claude-3's 1,271.

• Launched quietly on August 1, this experimental model has quickly gained attention for its superior performance.

• AI enthusiasts are eager about Gemini 1.5 Pro, though its future as the default model remains uncertain.

• Available for early testing in Google AI Studio, the Gemini API, and the LMSYS Chatbot Arena, its status as an experimental model suggests potential changes or withdrawal for safety or alignment reasons.

Why this important: Google's unexpected achievement with Gemini 1.5 Pro as the new AI benchmark leader might signal a significant shift in the competitive landscape, prompting major responses from industry rivals.

Prove safe and responsible use of AI - get $1,000 off Vanta

If you’re building or selling AI-powered products, demonstrating top-notch security practices and establishing trust is more important than ever.

With Vanta, you can quickly and easily demonstrate compliance with gold-standard AI frameworks like ISO 42001 and NIST AI RMF.

Vanta helps you prove secure deployment of AI, build customer trust, and accelerate your sales cycle.

Plus, with Vanta’s Questionnaire Automation, you can automate your responses to lengthy security questionnaires about your security posture and AI practices.

Learn more and claim a special offer of $1,000 off Vanta at the link below.

Get $1,000 off Vanta

Major Quick Developments

Meta has launched AI Studio, a platform powered by Llama 3.1, enabling users to build personalized AI chatbots for social media platforms like Instagram, Messenger, and WhatsApp.

Midjourney has released version 6.1 of its AI image generation platform, enhancing skin textures and text rendering for more realistic images and improved legibility — generally more beautiful.

Brett Adcock, CEO of Figure AI just teased — A brand new Figure O2 robot and it's fully upgraded.

Canva announced the acquisition of Leonardo AI, a startup known for its advanced image generative AI platform.

Stability AI has launched Stable Fast3D, a new model that generates 3D images in just half a second, achieving a 1200 times speed increase over previous models.

Friend, a US start-up, has launched a wearable AI device designed as a necklace that provides constant companionship by listening to the user and sending text messages based on their daily interactions.

Meta introduced Segment Anything Model 2 (SAM 2) builds on its predecessor with improved precision and expanded functionality for segmenting objects in both images and videos

BLACK FOREST LABS

📺 FLUX.1: A New AI Image Generation Contender

AI Researchs: Black Forest Labs just launched FLUX.1, a suite of AI image generation models that rival current leaders like Midjourney and DALL-E 3. It includes three variants: [pro], [dev], and [schnell].

The Details:

• The [pro] version offers top-quality performance and is available via API and for free on Replicate.

• The [dev] variant is an open-weight, non-commercial model that matches the [pro]'s quality while being more efficient.

• FLUX.1 [schnell] is an ultra-efficient, 4-step model for local development or personal use.

• Upcoming text-to-video generation model teased by Black Forest Labs, expected to rival Sora in quality.

Why this is important: FLUX.1's high quality and open-source options are set to democratize AI image generation, and with a potential leading video model on the way, the competition may face significant challenges ahead.

OPENAI VS MICROSOFT

🫣 Microsoft and OpenAI: New Frenemies in AI

AI Researchs: Microsoft has identified OpenAI, its long-term partner, as a competitor in its latest annual report.

The Details:

• Microsoft, which invested $13 billion in OpenAI, now views it as a competitor. OpenAI’s new search engine, SearchGPT, is a big reason for this shift.

• Despite being competitors, Microsoft still works with OpenAI—integrating its AI models into Azure and Bing. This competitive stance was part of their initial agreement.

• The relationship has had its bumps, like when Microsoft’s CEO wasn’t told about OpenAI’s brief ousting of Sam Altman. Adding to the mix, Microsoft recently brought on Mustafa Suleyman to head its AI division.

Why this is important: Microsoft and OpenAI's rivalry marks a big shift in the tech industry—potentially shaking up future AI advancements and market dynamics.

AI TUTORIAL

📽️ 5 Steps to Use Runway's Gen-3 Alpha Model

AI Researchs: Runway introduces Gen-3 Alpha, a new video generation model with enhanced fidelity, consistency, and motion, powered by multimodal training and advanced control features.

step-by-step:

1. Create Your Account: Sign up for a free account on the Runway website to get started.

2. Access Gen-3 Alpha: Log in and navigate to "Text/Image to Video" under "Runway's AI Tools" on your dashboard.

3. Upload Your Image: Select "Gen-3 Alpha" from the dropdown menu and upload your starting image by dragging and dropping or browsing your files.

4. Describe Your Video: Enter detailed instructions in the provided text box and select the video duration (5 or 10 seconds).

5. Generate and Review: Click "Generate" and review the resulting video, making adjustments as needed

AI Research Roundup

Researchers at the University of the Republic in Uruguay have developed an AI system, "Deep-TEMPEST," that intercepts and reconstructs HDMI signal emissions into readable screen content.

A fully-automatic robotic dentist has successfully completed its first human procedure, performing a dental crown preparation eight times faster than a human dentist, using advanced 3D imaging and AI for precise and efficient treatment.

AI Pointers

Intel CEO Pat Gelsinger says AI could use up to 30% of the nation's energy by 2030 as AI begins to permeate every aspect of our lives

Zapier co-founder Mike Knoop says AI language models have stalled in the progress to AGI and increasing scale will not help what is an inherently limited technology.

Hedge fund Elliott Management says Nvidia’s stock is in a bubble, claiming AI is overhyped and many of its applications won’t work as promised.

IBM has announced a quantum computer capable of solving complex problems in minutes, which would take today's supercomputers millions of years, pushing the boundaries of science and engineering.

• ✅ EduWiz: Generate documents in seconds

• ⚕️ Hamming Prompt Optimizer: A tool to automate 90% of manual prompt engineering using self-improving prompt optimizer.

• 🎧 Rome AI: An AI platform that creates podcasts on any topic by researching, breaking down subtopics, and crafting episodes for on-the-go listening.

• 🤖 Folderr: Build AI for any task, from a custom chat assistant using your data to a robust business workflow automation.

• 📞 toby: Real-time speech translation for any video call

• ⚙️ LangGraph Studio: AI-powered IDE for LLM development

• 🚦 Not Diamond: Use the world’s most advanced AI model router to select the best model for each task.

Vimeo's new AI-powered video translation tool uses generative AI to translate video content into multiple languages, maintaining the original speaker's voice and making global communication easier and more cost-effective.

Move AI's new motion capture app allows for 3D animation using a single camera for Game Development, eliminating the need for suits or markers.

Runway introduces Gen-3 Alpha Turbo, a faster version 7 times faster than the original Gen-3 Alpha, producing a 10-second video in 11 seconds.

Google Chrome is introducing new AI-powered features for desktop users, allowing enhanced browsing with Google AI and Gemini models.

OpenAI introduced the GPT-4o Long Output model, an extension of its GPT-4o, now allowing up to 64,000 tokens of output

OpenAI is gradually rolling out a new advanced voice mode for ChatGPT to a limited number of subscribers of ChatGPT Plus.

Taco Bell plans to use AI in drive-thru lanes at hundreds of locations across the U.S. by the end of 2024

Share Your Opinion