Cancer treatment breakthrough

Reading time: 7 minutes

Newsletter | News | Tools

Today we will discuss-

• 🚀Snowflake and Nvidia open up new possibilities

• 🚶AI helps speed up cancer radiotherapy

• 🤼‍♂️DeepMind's chatbot to challenge ChatGPT

All this and more - Let's dive in!

🍚What’s cooking in the newsroom

Snowflake and Nvidia join forces to make AI more accessible

📰News - Snowflake and Nvidia are partnering to make it easier for businesses to build and deploy AI models using their own data. Snowflake is a cloud data analytics company that provides a platform for businesses to store and analyze their data. Nvidia is a computing company that specializes in AI and graphics processing units.

Nvidia is embedding its NeMo platform into Snowflake Data Cloud, which means that NeMo will be available to Snowflake customers as a service. NeMo is a platform for training and running generative AI models.

🤔How would businesses benefit from this move? A retail company could use the partnership to build an AI model that can personalize product recommendations for customers. The model would be trained on data from the company's purchase history, and it could be used to suggest products that customers are likely to be interested in. This could help the company to increase sales and improve customer satisfaction.

🤓Will Nvidia benefit from this move? Nvidia's GPUs are used to power AI workloads. By making it easier for businesses to use generative AI models, Nvidia will increase the demand for its GPUs. This will help Nvidia to grow its sales and revenue.

AI-powered radiotherapy could shorten cancer treatment times

🗞️News - A new AI-powered program that can help cancer patients start radiotherapy sooner is to be offered to all NHS trusts in England.

Well, the NHS Artificial Intelligence Laboratory (NHS AI Lab) is a national research and development center that is focused on using artificial intelligence to improve healthcare in England.

🤔How was this AI developed? The program, called InnerEye, was developed by researchers at Addenbrooke's Hospital in Cambridge. It uses AI to analyze medical images of tumors and identify the best areas to target with radiotherapy. This can help to speed up the treatment process and reduce the time patients have to wait to start treatment.

😵What’s more? The government has been funding AI projects in the NHS for a while, but this is the first NHS-developed AI program that has been released as a medical imaging device.

Doctors would still check each of the contours drawn by the AI program, but the researchers say it is about 90% accurate, with clinicians approving its work without any corrections about two-thirds of the time.

DeepMind's new chatbot aims to take on ChatGPT

🖨️News - Demis Hassabis, CEO of Google DeepMind, says the company is working on a new chatbot called Gemini that he believes will surpass ChatGPT, a rival chatbot developed by OpenAI.

Demis also revealed that his engineers are using techniques from AlphaGo.

😎But how would the AlphaGO technique help Gemini?

• AlphaGo was a computer program developed by Google DeepMind that defeated a human world champion in the game of Go in 2016. It was able to do this by using a technique called reinforcement learning from human feedback (RLHF), which allows AI systems to learn from experience and improve their performance over time.

• Now, Demis’ team is planning to use the RLHF technique to train Gemini. The chatbot will, basically, be trained on a massive dataset of text and code, which will allow it to generate more creative and informative text than ChatGPT.

Gemini is still under development and could take a number of months to be completed. It is costing DeepMind tens or hundreds of millions of dollars.

🧐What else is happening?

• Amazon's new robots are rolling out an automation revolution

• Microsoft AI introduces ZeRO++ for efficient large-model training

• Thomson Reuters acquires legal tech startup Casetext for $650M in cash

• AI startup CalypsoAI raises $23 million to add guardrails to generative AI models

• Ramp acquires customer support startup Cohere.io

🤖Technical jargon simplified

• Data Preprocessing: It refers to the steps taken to clean, transform, and prepare data for AI algorithms. The process involves removing noise, handling missing values, normalizing data, and ensuring it is in a suitable format for analysis.

• Overfitting: It occurs when a machine learning model becomes too complex and starts to memorize the training data instead of learning generalizable patterns. It leads to poor performance on new, unseen data.

• Underfitting happens when a machine learning model is too simple and fails to capture the underlying patterns in the data. It results in low accuracy or performance on both training and new data. Finding the right balance is important for optimal model performance.

 🤯Discover mind-blowing AI tools

1. Deal Code - Helps sales professionals to compose highly personalized messages to potential customers quickly

2. UnlimitedBG - Tool that can automatically remove image backgrounds in seconds with professional-quality results

3. Codeium - Enables developers to quickly make changes to unfamiliar languages

4. Monster API - A platform that provides scalable and affordable APIs for various generative AI models

5. Protection Guard - Enforces structure and types on the output of language models, making them more efficient and compliant with regulations

6. Rows AI - An all-in-one spreadsheet tool that uses AI to analyze, summarize and transform data

7. VOGE - A revolutionary app that creates hyper-realistic AI avatars in just a few minutes

8. Torq - An AI-powered productivity assistant that helps users streamline their email communication

9. Magic Studio - A suite of tools to help create amazing product photos and stunning profile pictures

10. FigCopy - A tool that helps UI designers generate realistic and engaging copy for their designs

✨Mythbuster

Myth - Neural networks are like the human brain

Fact - The myth that neural networks are like the human brain is not entirely accurate. While neural networks are inspired by the way that the brain works, they are much simpler and do not fully replicate the complexity of the brain.