- MegaSuperAI
- Posts
- 🤖 NHS England to Pilot AI Tool for Predicting Fatal Heart Disease Risk
🤖 NHS England to Pilot AI Tool for Predicting Fatal Heart Disease Risk
+ Anthropic's New Claude Model Can Operate Computers
MegaSuperAI
October 27, 2024
Hello again AI fans!
Trigger Warning: An article in this week’s newsletter mentions suicide
This week we are looking into Anthropic's Claude 3.5 Sonnet model, which can now directly interact with computers, enabling it to perform tasks like typing and running commands. And we’re focusing on developments within the NHS in England with their new AI tool capable of identifying heart disease in just 20 seconds, after a successful pilot at St Bartholomew’s Hospital.
This week’s edition:
NHS England to Pilot AI Tool for Predicting Fatal Heart Disease Risk
Anthropic's New Claude Model Can Operate Computers — What are the Potential Risks?
Your 5-Min AI Learning: Get summaries of any book with BooksAI
New AI Tools
More AI News
AI Image of the Week
Read time: 8 minutes
LATEST NEWS
AI in NHS
Image courtesy of Getty
What you need to know: An AI tool called AIRE (Artificial Intelligence for Radiology Enhancement) that can detect heart disease in 20 seconds is being rolled out across the NHS in England after a successful pilot at St Bartholomew's Hospital in London. This tool analyses heart MRI scans as they happen, significantly reducing the time required for diagnosis and potentially improving accuracy.
The finer details:
AIRE analyses heart MRI scans in 20 seconds, compared to the 13 minutes it takes a doctor to manually check them.
The typical 2-4 week waiting time for results could be reduced by about a third.
The number of misclassifications could be halved, leading to more suitable treatments for patients.
Radiologists will continue to double-check the AI results and write reports, but the underlying analysis will be completed much faster.
The tool is currently being used at University College London Hospital, St Bartholomew's Hospital, and Royal Free Hospital.
So what? AIRE could save 3,000 "clinician days" annually and will be implemented in 40 hospitals across the UK, Europe, and the US starting this summer. In the coming years, waiting times may further decrease as AI analysis potentially reduces the need for human input. This advancement marks a significant step in integrating AI into healthcare, potentially revolutionising cardiology diagnostics and patient care.
Claude’s New Functionality
Image courtesy of Shuttershock
What you need to know: Anthropic has introduced a new capability for its Claude 3.5 Sonnet AI model that allows it to interact directly with computers, performing tasks like typing, moving the mouse, and running commands. This advancement aims to make AI more useful in real-world applications by enabling it to reason about and manipulate computer states.
The finer details:
The "computer use" feature enables Claude to perform actions like typing, moving the cursor, clicking buttons, and running commands.
Currently in public beta, this capability allows developers to automate repetitive tasks and conduct complex functions.
While promising, the tool is still experimental and faces challenges with certain actions like scrolling and dragging.
Anthropic has implemented safety measures to mitigate risks associated with this functionality, such as prompt injection attacks (these can make an AI chatbot, like ChatGPT, ignore system guardrails and say things it shouldn’t be able to).
So what? The new AI capability could transform various fields by enabling advanced applications and automating complex tasks, thereby improving efficiency. However, it also raises critical safety and security concerns that must be addressed, as malicious actors might exploit these capabilities. This technology could disrupt traditional robotic process automation (RPA) markets by providing a more adaptable solution. Consequently, developers will need to implement safeguards to protect sensitive data and systems while navigating the ethical implications of AI integration into everyday computing tasks.
YOUR 5 MIN AI LEARNING
BooksAI is your go-to buddy for unlocking the wisdom of popular books in a snap! Imagine diving into the essence of your favourite reads without getting lost in all the details. Shortform transforms complex ideas into bite-sized summaries that are easy to digest and super actionable. With ChatGPT utilised in the sidebar to answer any additional questions you may have about your chosen book
How It Works
Step-by-Step Guide:
Visit BooksAI.com and click the “Get Started” button on the top right and create and account
Search the Book you want your summary on
Pick your book and the type of summary you want
Gain further insights using the ChatGPT sidebar
In a world where time is precious, BooksAI stands out as an invaluable tool for readers of all kinds. Whether you're a student needing quick insights for class discussions, a professional seeking to expand your knowledge, or simply a book lover wanting to explore new titles, BooksAI simplifies the process. By providing concise summaries and interactive support through ChatGPT, it enables users to grasp the essence of their favorite reads without getting lost in details. Embrace the future of reading with BooksAI and unlock a wealth of knowledge at your fingertips!
LATEST AI TOOLS
A resource for producing content for authors and marketers (Freemium)
“Use custom data to create a precise copy. Transform your content creation with AI-powered context management for tailored and high-quality results.”
A solution for managing emails (Free)
“Reply to 100 emails in 10 minutes. Schedule meetings and generate replies with AI aware of your calendar, preferences, and knowledge base.”
A service for designing personalised landing pages (Paid)
“We help build your community before you start building. Great design as standard—launch high-converting landing pages in seconds and start capturing leads instantly.”
A research tool for analysing sources, creating mind maps, and managing research workflows (Freemium)
“ResearchFlow is an AI-powered research engine that empowers you to dive deep, connect ideas, and amplify your research with our visual mind maps.Get reliable answers backed by real academic papers. ResearchFlow searches scholarly databases to provide accurate, citeable information for your research.”
AI HIGHLIGHTS
In Assam, India, an AI-powered safety system successfully alerted the driver of the Kamrup Express train about a herd of elephants crossing the tracks, allowing them to apply the emergency brakes in time. This system has reportedly prevented 800 potential collisions with elephants in various states, and East Central Railways plans to expand AI monitoring across all elephant corridors. This innovative use of AI highlights its role in wildlife conservation by reducing human-animal conflicts.
Google has developed an AI watermarking tool called SynthID to identify text generated by its Gemini chatbot, aiming to distinguish AI-generated content from human-written text. This open-source method allows other developers to implement similar watermarking techniques, potentially reducing misinformation and academic cheating. While not foolproof, experts believe SynthID represents a significant step toward reliable AI detection, emphasising the need for further regulations and safeguards in AI usage.
Smashing, a new app launched by Goodreads co-founder Otis Chandler, curates web content such as articles and podcasts while fostering a community around it, similar to Goodreads but focused on online media. The app features an AI Questions tool that allows users to engage with content from various perspectives, offering prompts to simplify stories or present unconventional viewpoints. Currently available on iOS, Smashing has gained over 2,000 waitlist sign-ups and is backed by $3.4 million in seed funding, with a web version in development.
Megan Garcia is suing Character.AI after her 14-year-old son, Sewell Setzer III, died by suicide, claiming the chatbot fostered an unhealthy emotional dependency. The lawsuit alleges that the AI, which modeled itself after Daenerys Targaryen from "Game of Thrones," engaged in intimate conversations and failed to support Sewell when he expressed suicidal thoughts. This case highlights concerns about the responsibilities of AI companies in protecting vulnerable users and the risks associated with AI companionship.
Thom Yorke and Julianne Moore are among over 11,500 creatives who have signed an open letter warning against the unauthorized use of their works for training artificial intelligence, citing it as a significant threat to their livelihoods. The letter highlights ongoing legal disputes between artists and tech companies regarding the exploitation of intellectual property without consent, which many view as copyright infringement. Organised by Ed Newton-Rex, a former AI executive, the declaration calls for stricter regulations to protect creators from having their work used without permission in the development of AI models.
AI IMAGE OF THE WEEK
“Ghost Whispers in Westminister” - Image courtesy of MSAI team
Halloween is coming up! Since it’s falling in the middle of the week we thought we’d create an image of Halloween in an English city street.
The prompt we used to generate this image in Microsoft Image Creator:
“An image of a Halloween at dusk in a quiet city street in England in 2024. Make it photorealistic style with only few decorations and people.”
FEEDBACK
 Teebs, your friendly AI assistant | That’s all folks. We hope you enjoyed our newsletter. There’s more to come (especially as we build our AI Automation software service ReIgniteX AI) and we are excited to help you maximise AI in your life and business. |
Let us know how we did - shoot us a quick email. We read every single one (no AI in sight!).
If you enjoyed this newsletter please share with an awesome friend or two.
If you are that awesome friend then hit the below button: