Kala

Italy's privacy regulator has banned OpenAI's ChatGPT for collecting and storing Italian user's personal data without any legal basis. OpenAI has also been accused of violating the EU General Data Protection Regulation and an investigation has been launched.

The ban comes after an open letter was published on Twitter, calling for a six-month pause on the development of the system due to potential risks to society. OpenAI must comply with the ban with immediate effect or face a fine of up to €20 million.

Pope Francis praises the benefits of technology and AI for human family, but calls for responsible and ethical use to promote human dignity and the common good.

He appreciates the creativity and nobility of human participation in AI technology development, but urges for a consistent commitment to acting ethically and responsibly.

Check why Cloudways is better than other hosting providers and start your free trial now - no credit card required!

ChatGPT had to be taken offline early this week because of a bug in an open source library that allowed users to see the chat history of other active users.

ChatGPT Plus subscribers may have had their personal information, including their name, email address, payment information, and last four digits of their credit card exposed.

The bug has been patched, and technical details will be published after an investigation.

The article discusses the use of orchestrators in automating machine learning processes and ensuring the reproducibility of results.

The three main orchestrators discussed are Kubeflow Pipelines, Airflow, and Prefect. The article compares the advantages and disadvantages of each orchestrator and concludes that the choice of orchestrator depends on the specific project requirements.

Kubeflow was the preferred choice for the authors due to its CI/CD configuration and containerization advantages. However, Prefect is recommended for small projects or teams with a limited budget, and Airflow is ideal for running a variety of tasks using a mature and developed tool.

Learn how to speed up training of PyTorch and Hugging Face models using a reduction server through Vertex AI feature, optimizing bandwidth and latency to distribute training across multiple Nvidia GPUs.

The all-reduce based algorithm synchronously averages gradients across multiple devices, reducing training time and costs.

The post dives into a detailed description of datum parallelism and reduction server's architecture and impact on training throughput.

>>Check out the accompanying notebook and video to learn more!

Shopify has enhanced its machine learning platform, Merlin, to include an end-to-end machine learning workflow. Merlin Online Inference is a flexible, robust, and low-latency solution that enables Shopify to deploy machine learning models in real-time at scale.

With the growing importance of close user-face product integration, it has become necessary to utilize machine learning models with low latency, which poses many different use cases and requirements.

Shopify uses its internal service ecosystem to make sure the service deployment is scalable and that the Merlin Online Inference service's health is observed.

Buzzfeed shares its experience in incorporating generative AI product (like OpenAI's GPT) last year.

Buzzfeed says AI will bring a new era of creativity that opens up fresh content formats and novel ways for content creators to interact with audiences.

Buzzfeed staff used OpenAI playground to understand how language models work and to gain skills in creating an effective prompt. Through this, they learned moderation is essential to building guardrails around the prompt.

OpenAI has released its latest language model, GPT-4, which has several new features that set it apart from previous versions like ChatGPT and GPT-3.5.

The five biggest differences are that GPT-4:

• can understand and process images,
• is harder to trick,
• has a longer memory,
• is more multilingual,
• has different "personalities" that can be customized to suit specific needs.

GPT-4 is expected to be available to developers soon via API and will be featured on OpenAI's paid service, ChatGPT Plus.

👉 Get an immediate 20% off on all products using the code "THANKSFAUN"

Checkout "OpenAI GPT For Python Developers" by Aymen El Amri (@eon01), founder of FAUN.

In this practical guide, you will build intelligent real-world applications using GPT models, DALL-E, Whisper, CLIP, and more tools from the OpenAI and ML ecosystem.

Rest assured, you don't need to be a Python expert, a data scientist, or a machine learning engineer to follow this guide. All you need is a basic understanding of how Python works and the openness to learn new concepts!

By the end of your learning journey, you will have built applications such as:

• A fine-tuned medical chatbot assistant
• An intelligent coffee recommendation system
• An intelligent conversational system with memory and context
• An AI voice assistant like Alexa but smarter
• A Chatbot assistant to help with Linux commands
• A semantic search engine
• A news category prediction system
• An image recognition intelligent system (image to text)
• An image generator (text to image)
• and more!

Join the Web3 Revolution: Status.im is seeking a talented Infrastructure Engineer!

Contribute to the future of secure, private, and open web3 with Ethereum-based tools, DApps, and cutting-edge blockchain technology. Work remotely with a diverse, global team.

Apply now and help build the decentralized web of tomorrow.

Join Greg Brockman, President and Co-Founder of OpenAI, for a developer demo showcasing GPT-4 and some of its capabilities/limitations.

Advanced relational operations for pandas DataFrames

Next-Gen GPT-ize Intelligent Application Architecture - A prompt engineering approach with LLM AI Dev Framework

Giving Kubernetes SRE superpowers to everyone

Simple Python package to preview and develop streamlit apps in jupyter notebooks