AWS Amazon offers cloud adoption framework for AI/ML

What's Happening This Week in AI

Author

Anne Borsohalmi
July 28, 2023

Happy Friday! In this edition, we bring you a diverse range of insights and updates from the world of artificial intelligence, machine learning, natural language processing, regulations, research, and more.

We hope you find this newsletter informative and insightful as you navigate the dynamic world of AI and its applications across industries.

Please subscribe to The AI Collective Word today to receive your free newsletter directly in your inbox and join our ever-growing network.

I hope you enjoy this week’s newsletter!

Share it with a friend or colleague if you find it helpful.

ML and NLP

Created by AWS Amazon, CAF-AI is a cloud adoption framework for AI/ML transformation on AWS. It helps you define and execute an AI strategy that aligns with your business outcomes. It guides you through an iterative process of assessing, deriving, and implementing action items across different domains and capabilities.

Stability AI and its CarperAI lab have introduced two new open access Large Language Models (LLMs) called FreeWilly1 and FreeWilly2. These models exhibit impressive reasoning capabilities across various benchmarks. The training for the FreeWilly models stems from the Microsoft methodology.

FreeWilly1 is based on the LLaMA 65B foundation model, and FreeWilly2 utilizes the LLaMA 2 70B foundation model and delivers a performance comparable to GPT-3.5 for certain tasks.

Both models serve as research experiments and have been released to encourage open research under a non-commercial license.

The paper discusses the challenge of creating an ideal long-term memory mechanism for Large Language Model (LLM) based chatbots, enabling continual learning and complex reasoning. The proposed RecallM architecture aims to achieve adaptable and updatable long-term memory for AGI systems. The architecture demonstrates improved temporal understanding, addressing the limitations of vector databases in belief updating and temporal relations while providing an effective solution for question answering over large text corpora. The goal is to enhance LLMs to achieve true Artificial General Intelligence (AGI).

TRENDING NEWS

Created in Canva

Partnerships are key to positioning organizations to excel in the AI space and to help organizations with their IT strategy. Here are recent key partnerships:

  • Softbox is teaming up with Symbotic to build AI-powered warehouses. - Reuters

  • Box extends AI efforts with Microsoft 365 Copilot integration - VentureBeat

  • ServiceNow partners with Nvidia and Accenture on ‘AI Lighthouse’ for rapid enterprise AI adoption - VentureBeat

  • Genpact teams up with Microsoft to empower its workforce with generative AI tools - VentureBeat

  • Google Cloud and CareCloud are working together to expand small- and medium-sized healthcare providers’ generative artificial intelligence capabilities. - Health IT Analytics

INDUSTRY TRENDS

A study by ClearML and AIIA surveyed F-1000 C-level executives about generative AI adoption. It found that most executives face budget and resource constraints, and cannot measure the ROI of their AI/ML projects. The study suggests that more investment in AI is needed to create value and innovation.

An AI spy van caught hundreds of drivers using phones or not wearing seat belts. The van used two cameras to detect the offenses and verified them with human reviewers. The police said the technology revealed how common the offenses were after a week of surveillance. The AI technology will be used as a reminder to people that distracted driving kills, and these offenses will be punished.

REGULATIONS

Created in Canva

The EU AI Act is a proposed law that aims to regulate artificial intelligence in Europe. A group of open-source AI organizations has published a policy paper to advocate for open and transparent AI development practices. They warn that the AI Act could favor closed and proprietary AI models and harm the open AI ecosystem.

Four leading AI companies have formed the Frontier Model Forum, a self-regulatory body to set safety standards for advanced AI systems. The forum aims to foster AI safety research and technical evaluations, as well as information sharing among companies and governments. The initiative follows the companies' pledges to the White House to test their systems and flag AI-generated content.

PLATFORM ENGINEERING

Generative AI models are powerful tools for creating natural language responses from data. However, they are not the only option for enterprises that want to use AI for specific tasks. Some startups offer open source AI models that can be customized and trained with enterprise data, giving more control and flexibility. These models can help enterprises leverage their data to provide information or assistance to customers or employees in various domains, such as healthcare, finance, education and government. Generative AI is still an emerging technology that has many potential applications for improving business processes.

Microsoft has launched a new program called AIM (Accelerate Innovate Move) to assist on-premises Dynamics 365 customers with cloud migrations. The program offers tools and services for Dynamics AX, Dynamics CRM, Dynamics GP, Dynamics NAV, Dynamics SL, and Dynamics 365 Business Central customers to transition their critical business processes to the cloud. Microsoft emphasizes the importance of embracing cloud technologies, particularly generative AI, to stay competitive in the evolving business landscape. The AIM program includes cloud migration assessment, access to Microsoft services and support, and licensing offers for incremental migration of critical business processes.

RESEARCH

Cortical Labs researchers receive a $600k grant from Australia’s Office of National Defense to merge brain cells with AI.

Artificial intelligence (AI) and machine learning (ML) have the potential to positively impact anesthesiology by improving patient safety and outcomes. These technologies can help capture and analyze the massive amounts of data generated by anesthesia patients, leading to better intraoperative practices and postoperative outcomes. AI can assist anesthesiology professionals by reducing cognitive load, allowing them to focus on critical aspects of patient care while utilizing AI as a tool for navigating and interpreting data. However, it is important to strike a balance between technology and the human connection in anesthesia, as these tools should free up time for providers to be physically and emotionally present for patients. Care teams need to identify high-value use cases for AI and integrate it effectively into clinical workflows to improve patient care.

The study showed how AI and synthetic biology can fight infectious diseases in three ways: finding new drugs, understanding infection biology, and improving diagnostics. AI and synthetic biology can work together to search for new drugs, analyze complex factors of infection, and detect diseases faster and more accurately. The study highlighted the potential of AI and synthetic biology to address the public health challenge of infectious diseases.

USE CASES

Emergency medical services (EMS) are evaluating the usefulness of artificial intelligence (AI) in improving emergency care efficiency and patient outcomes. AI and health IT tools can help EMS with uptime and stability of mobile incident management software, as well as enhancing patient triage efficiency. The University of Pittsburgh Medical Center (UPMC) EMS department is developing an ML tool that will use electrocardiograms (ECG/EKG) to classify cardiac events and improve triage methods. By processing large amounts of raw data from ECG readings, the machine learning tool can provide a more comprehensive view of the patient's heart health to accurately identify potential cardiac events.

RISK AND SECURITY MANAGEMENT

Splunk, a provider of machine data indexing and search technology, has launched six AI-driven automation tools for security operations teams. The company's threat research team trained deep learning AI models on security data to create specific tools that can detect security threats such as DNS exfiltration attempts. Splunk's artificial intelligence and machine learning technologies are available through its existing search processing language, machine learning toolkit and app for data science and deep learning. These tools, which can generate Notable Events workflows within the Splunk Enterprise Security security information and event management product, are capable of detecting security issues before they occur.

The SEC has adopted new rules on cybersecurity and AI conflicts of interest. Companies must report material cyber breaches within four days, unless law enforcement needs more time. They must also disclose their cyber risk management practices regularly.

Sysdig Sage AI is a new feature that uses LLMs to improve cloud security and monitoring. It can perform complex reasoning and actions to help users prevent and respond to incidents. Sysdig Sage AI is not a simple LLM wrapper, but a controller that integrates different LLMs with security domain knowledge.

ETHICS

Ryanair uses facial recognition to check customers' identities when they book through online travel agents. NOYB, a digital rights group, filed a complaint against Ryanair in Spain, saying this violates data protection rights. Ryanair says it needs to verify identity for safety and security reasons and offers other options that are less convenient.

Resources

Learn to program in TensorFlow Lite for microcontrollers so that you can write the code, and deploy your model to your very own tiny microcontroller. Self-paced course, Duration: 5 weeks (2 to 4 hours per week)

Learn the basics of computer science from Harvard University. This is CS50, an introduction to the intellectual enterprises of computer science and the art of programming. Duration: 24 hours of lectures.