🚨 The Godfather of AI leaves Google and Warns of Danger Ahead..
On Geoffrey Hinton leaving Google, HuggingFace new ChatGPT Rival, and Stability AI's new DeepFloyd IF. Enjoy!
Welcome to this week's edition of AlphaSignal the newsletter for AI experts. Whether you are a researcher, engineer, developer, or data scientist, our summaries ensure you're always up-to-date with the latest breakthroughs in AI.
Let's get into it!
Here's what’s on the menu today:
Top Announcements: Ex-DeepMind new LLM, IBM cutting jobs, and more.
Geoffrey Hinton leaves google.
HuggingFace’s new ChatGPT rival.
Top of Reddit: Voice cloning, Upscaling old games, and more.
Stability AI releases a new text-to-image model.
📑 RELEASES & ANNOUNCEMENT
1.LinkedIn And DeepMind’s Co-Founders Launch ChatGPT Rival ‘Pi’ With Higher Emotional Intelligence ↗
Unlike other chatbots, Pi is designed to be personal, and while it can provide fact-based answers, it avoids fostering parasocial relationships. Inflection intends to enhance Pi with real-time content, share links, and eventually incorporate a user's calendar, email, and other documents to better manage their time. The company has raised $225 million in funding to date. Give Pi a try for yourself.
2. IBM is pausing hiring on jobs that could be replaced by AI ↗
IBM's CEO, Arvind Krishna, has announced plans to pause hiring for back-office functions, including human resources, which he believes could be replaced by AI and automation over the next five years. This could lead to the loss of approximately 7,800 jobs. While mundane tasks such as employment verification and employee transfers will likely be fully automated, evaluating workforce composition and productivity may not be replaced over the next decade.
3. Segment Anything for Videos? Track-Anything is a flexible and interactive tool for video object tracking and segmentation ↗
Track-Anything applies SAM, XMem, and E2FGVI to videos, providing video inpainting capabilities, as well as video and multi-object tracking with segmentation masks.
4. Latest NVIDIA Graphics Research Advances Generative AI’s Next Frontier ↗
NVIDIA will showcase around 20 research papers at SIGGRAPH 2023. The research covers generative AI and neural graphics, including the development of generative AI models that turn text into personalized images, inverse rendering tools that transform still images into 3D objects, neural physics models that simulate complex 3D elements, and neural rendering models that generate real-time, AI-powered visual details.
5. Runway launches Gen-1 iPhone app ↗
AI startup Runway has launched its first mobile app for iOS, which features its video-to-video generative AI model Gen-1. Users can record a video from their phones and transform it into an AI video in minutes, or use text prompts, images or style presets to alter existing videos.
The Buyer’s Guide to Evaluating ML Feature Stores & Feature Platforms
If you’re looking to adopt a feature store or platform for machine learning and don’t know where or how to start your research, then this guide is for you.
Download the free guide to:
Access a comprehensive framework for understanding the capabilities of different feature stores and platforms
Get tips to on how to use a data-driven approach to evaluate vendors
Learn how the right solution can improve ML model accuracy
Geoffrey Hinton, the 'Godfather of AI', Leaves Google and Warns of Danger Ahead
Geoffrey Hinton, the renowned AI pioneer, recently quit his position at Google to focus on ethical AI development and speak openly about the risks associated with AI. Hinton is one of the three “Godfathers of AI” who won the Turing Award in 2018 for their groundbreaking work on neural networks.
In a recent interview with The New York Times, Hinton expressed his concerns about the potential implications of rapid advancements in AI, including the existential risk posed by AI systems becoming more intelligent than humans. “I console myself with the normal excuse: If I hadn’t done it, somebody else would have,” Dr. Hinton said during a lengthy interview last week in the dining room of his home in Toronto, a short walk from where he and his students made their breakthrough.
In light of his observations of new large language models like GPT-4, Hinton cautions about several key issues:
“Machines surpassing human intelligence: Hinton believes AI systems like GPT-4 are on track to be much smarter than initially anticipated, potentially possessing better learning algorithms than humans.
Risks of AI chatbots being exploited by “bad actors”: Hinton highlights the dangers of using intelligent chatbots to spread misinformation, manipulate electorates, and create powerful spambots.
Few-shot learning capabilities: AI models can learn new tasks with just a few examples, enabling machines to acquire new skills at a rate comparable to, or even surpass, that of humans.
Existential risk posed by AI systems: Hinton warns about scenarios in which AI systems create their own subgoals and strive for more power, surpassing human knowledge accumulation and sharing capabilities.
Impact on job markets: AI and automation can displace jobs in certain industries, with manufacturing, agriculture, and healthcare being particularly affected.”
Want to promote a product, job, or event to 100,000+ AI researchers and engineers? You can reach out to us here.
Hugging Face launches open-source version of ChatGPT
Hugging Face, a well-known player in the open-source AI development arena, has unveiled open-source 30B chatbot alternative to ChatGPT, named HuggingChat. The new platform boasts a user interface that lets users interact with an open-source chat assistant called Open Assistant, created by the nonprofit organization LAION. HuggingChat, which has a web interface for testing and an API for integration with third-party apps and services, can write code, emails, and even rap lyrics, just like ChatGPT. The emergence of HuggingChat, which is still in its infancy, is a significant milestone for Hugging Face, as it reaffirms their commitment to open-source development and accessibility. However, HuggingChat's economic viability remains to be seen, as licensing issues still need to be resolved.
From an AI industry perspective, HuggingChat's arrival carries significant implications. Hugging Face's CEO, Clem Delangue, has stated that "we need open-source alternatives to ChatGPT for more transparency, inclusivity, accountability, and distribution of power." HuggingChat's open-source platform provides transparency, accessibility, and inclusivity that closed-source models lack, paving the way for a more democratic AI landscape. In addition, HuggingChat's potential for third-party app integration has sparked talk of it becoming the Android App Store equivalent. This would mean that HuggingChat could level the playing field for smaller AI developers and reduce the monopoly of larger companies in the industry.
🤖 TOP OF REDDIT
This is surreal: ElevenLabs AI can now clone the voice of someone that speaks English (BBC's David Attenborough in this case) and let them say things in a language, they don't speak, like German.
Stability AI releases DeepFloyd IF, a powerful text-to-image model that can smartly integrate text into images
Stability AI has partnered with its AI research lab DeepFloyd to introduce the research version of its latest technology, called DeepFloyd IF.
This latest technology is designed to generate high-quality images from text inputs, and it boasts some impressive features. Notably, it uses the T5-XXL-1.1 language model as a text encoder to better understand text prompts, and it employs cross-attention layers to better align the text prompt and the generated image.
One standout feature is the model's ability to accurately apply text descriptions to generate images with various objects appearing in different spatial relations, a task that has been challenging for other text-to-image models. The model is also praised for its high degree of photorealism, with an impressive zero-shot FID score of 6.66 on the COCO dataset. In addition to text-to-image generation, the model offers zero-shot image-to-image translations, allowing for the modification of style, patterns, and details in the output image while maintaining the primary form of the source image. The model is initially released under a research license, with plans for a full open-source release in the future
Deep text prompt understanding
Application of text description into images
A high degree of photorealism
Aspect ratio shift
Zero-shot image-to-image translations