It Good To See The Real World Joining In And Deploying Clinically Relevant AI!

 This appeared last week:

Liverpool Hospital explores GenAI cardiology solution

Wednesday, 06 December, 2023

With cardiovascular disease responsible for one in four deaths across the nation, cardiologists play a critical role in Australia’s healthcare system. These specialists are challenged, however, by a constantly evolving field whose knowledge base is growing exponentially.

To address this challenge, Liverpool Hospital and Microsoft have collaborated on a generative artificial intelligence (GenAI) solution to assist cardiologists in surveying the vast volume of cardiology-related literature and guidelines that aids decision-making in clinical practice.

Their prototype, ‘The Cardiology Canon’, allows a clinician to interact with an extensive body of literature from the field of cardiology. Built using Microsoft Azure OpenAI Service, it was programmed on clinical trials from the past 40 years, along with guidelines from eminent cardiac societies in Australia, Europe and the United States.

The Microsoft Technology Centre team processed more than 500 individual documents and 80,000 pages to create a cognitive vector search index and made the solution accessible via a web application.

“We worked with Microsoft technologists to create a bespoke prototype to allow cardiologists and anyone training in the field to query vast volumes of literature to answer clinical questions or make nuanced comparisons, traversing the significant body of literature,” said Dr Hao Tran, Cardiology Advanced Trainee at Liverpool Hospital.

“The solution also successfully generated a comprehensive six-day training program on aortic stenosis — a common clinical problem among patients — and generated comprehensive learning resources in hours, rather than days.

“It was presented at the Liverpool Cardiology Academy academic seminar with Microsoft collaborators Viren Joseph, Leon Smith and Daniel Xu and was championed by senior Liverpool Hospital cardiologists Professor Craig Juergens and Dr Sidney Lo. It’s been a truly collaborative project that I believe will greatly benefit cardiologists and patients,” Tran added.

Dr Simon Kos, Chief Medical Officer at Microsoft Australia and New Zealand, said the collaboration demonstrated the invaluable contribution of hospital clinicians in the sphere of innovation, with their feedback iteratively helping to refine the solution’s large language model to suit clinical need.

“We’re excited to collaborate with Liverpool Hospital on this initiative and unlock the incredible potential of generative AI in health care. This innovative approach is empowering the hospital’s cardiologists to stay ahead in their field, ensuring the highest standard of patient care,” Kos said.

The announcement follows a new report by Microsoft and the Tech Council of Australia highlighting generative AI’s economic and productivity opportunities for Australia. Health care is identified as one of four key economic sectors poised to benefit from the technology, potentially contributing $5–13 billion in value annually to the sector in Australia by 2030.

According to the report, generative AI can automate 25% and augment 5% of nursing tasks, allowing nurses to spend more one-on-one time with patients. It also estimates that 30% of new drugs could be discovered using the technology by 2025.

Here is the link to the release:

https://www.hospitalhealth.com.au/content/technology/news/liverpool-hospital-explores-genai-cardiology-solution-397830128

Good it seems stuff is really happening!

David.

It Seems It Is The Time Of Year For Predictions For Next Year!

 This appeared a few days ago:

 Thursday, 07 December 2023 04:13

Twelve AI predictions for 2024

By David M Williams

In time for the 12 days of Christmas, data, AI, and analytics software company SAS has released its 12 predictions about AI for the coming year.

 Generative AI will augment (not replace) a comprehensive AI strategy

“Generative AI technology does a lot of things, but it can’t do everything. In 2024, organizations will pivot from viewing generative AI as a stand-alone technology to integrating it as a complement to industry-specific AI strategies. In banking, simulated data for stress testing and scenario analysis will help predict risks and prevent losses. In health care, that means the generation of individualized treatment plans. In manufacturing, generative AI can simulate production to identify improvements in quality, reliability, maintenance, energy efficiency and yield.”

– Bryan Harris, Chief Technology Officer, SAS

Note that earlier this year, SAS committed $1 billion to AI-powered industry solutions.

AI will create jobs

“In 2023, there was a lot of worry about the jobs that AI might eliminate. The conversation in 2024 will focus instead on the jobs AI will create. An obvious example is prompt engineering, which links a model's potential with its real-world application. AI helps workers at all skill levels and roles to be more effective and efficient. And while new AI technologies in 2024 and beyond may cause some short-term disruptions in the job market, they will spark many new jobs and new roles that will help drive economic growth.”

– Udo Sglavo, Vice President of Advanced Analytics, SAS

 AI will enhance responsible marketing

“As marketers, we must consciously practice responsible marketing. Facets of this are awareness of the fallibility of AI and alertness to possible bias creeping in. While AI offers the promise of enhanced marketing and advertising programs, we know that biased data and models beget biased results. In SAS Marketing, we are implementing model cards that are like an ingredient list, but for AI. Whether you create or apply AI, you are responsible for its impact. That's why all marketers, regardless of technical know-how, can review the model cards, validate that their algorithms are effective and fair, and adjust as needed.”

– Jennifer Chase, Chief Marketing Officer, SAS

Financial firms will embrace AI amid a Dark Age of Fraud

“Even as consumers signal increased fraud vigilance, generative AI and deepfake technology are helping fraudsters hone their multitrillion-dollar craft. Phishing messages are more polished. Imitation websites look stunningly legitimate. A crook can clone a voice with a few seconds of audio using simple online tools. We are entering the Dark Age of Fraud, where banks and credit unions will scramble to make up for lost time in AI adoption – incentivized, no doubt, by regulatory shifts forcing financial firms to assume greater liability for soaring APP [authorized push payment] scams and other frauds.”

– Stu Bradley, Senior Vice President of Risk, Fraud and Compliance Solutions, SAS

Shadow AI will challenge CIOs

“CIOs have struggled with ‘shadow IT’ in the past and will now confront ‘shadow AI’ – solutions used by or developed within an organization without official sanction or monitoring by IT. Well-intentioned employees will continue to use generative AI tools to increase productivity. And CIOs will wrestle daily with how much to embrace these generative AI tools and what guardrails should be put in place to safeguard their organizations from associated risks.”

– Jay Upchurch, Chief Information Officer, SAS

Multimodal AI and AI simulation will reach new frontiers

“The integration of text, images and audio into a single model is the next frontier of generative AI. Known as multimodal AI, it can process a diverse range of inputs simultaneously, enabling more context-aware applications for effective decision-making. An example of this will be the generation of 3D objects, environments and spatial data. This will have applications in augmented reality [AR], virtual reality [VR], and the simulation of complex physical systems such as digital twins.”

– Marinela Profi, AI/Generative AI Strategy Advisor, SAS

 Digital-twin adoption will accelerate

“Technologies like AI and IoT [Internet of Things] analytics drive important sectors of the economy, including manufacturing, energy and government. Workers on the factory floor and in the executive suite use these technologies to transform huge volumes of data into better, faster decisions. In 2024, the adoption of AI and IoT analytics will accelerate through the broader use of digital-twin technologies, which analyse real-time sensor and operational data and create duplicates of complex systems like factories, smart cities and energy grids. With digital twins, organizations can optimize operations, improve product quality, enhance safety, increase reliability and reduce emissions.”

– Jason Mann, Vice President of IoT, SAS

 Insurers will confront climate risk, aided by AI

“After decades of anticipation, climate change has transformed from speculative menace to genuine threat. Global insured losses from natural disasters surpassed $130 billion in 2022, and insurers worldwide are feeling the squeeze. US insurers, for example, are under scrutiny for raising premiums and withdrawing from hard-hit states like California and Florida, leaving tens of millions of consumers in the lurch. To survive this crisis, insurers will increasingly adopt AI to tap the potential of their immense data stores to shore up liquidity and be competitive. Beyond the gains they realize in dynamic premium pricing and risk assessment, AI will help them automate and enhance claims processing, fraud detection, customer service and more.”

– Troy Haines, Senior Vice President of Risk Research and Quantitative Solutions, SAS

AI importance will grow in government

“The workforce implications of AI will start being felt in government. Governments have a hard time attracting and retaining AI talent since experts command such high salaries, however, they will aggressively recruit for expertise to support regulatory actions. And like enterprises, governments will also increasingly turn to AI and analytics to boost productivity, automate menial tasks and mitigate that talent shortage.”

– Reggie Townsend, Vice President of the SAS Data Ethics Practice

 Generative AI will bolster patient care

“To advance health and improve patient and member experiences, organizations will further develop generative AI-powered tools in 2024 for personalized medicine, such as the creation of patient-specific avatars for use in clinical trials and the generation of individualized treatment plans. Additionally, we will see the emergence of generative AI-based systems for clinical decision support, delivering real-time guidance to payers, providers and pharmaceutical organizations.”

– Steve Kearney, Global Medical Director, SAS

Deliberate AI deployment will make or break insurers

“In 2024, one of the top 100 global insurers will go out of business as a consequence of deploying generative AI too quickly. Right now, insurers are rolling out autonomous systems at breakneck speed with no tailoring to their business models. They’re hoping that using AI to crunch through claims quickly will offset the last few years of poor business results. But after 2023’s layoffs, remaining staff will be spread too thin to enact the necessary oversight to deploy AI ethically and at scale. The myth of AI as a cure-all will trigger tens of thousands of faulty business decisions that will lead to a corporate collapse, which may irreparably damage consumer and regulator trust.”

– Franklin Manchester, Global Insurance Strategic Advisor, SAS

Public health will get an AI boost from academia

“Public health is achieving technological modernisation at an unprecedented rate. Whether overdoses or flu surveillance, using data to anticipate public health interventions is essential. Forecasting and modelling are rapidly becoming the cornerstone of public health work, but government needs help. Enter academia. We will see an increase in academic researchers carrying out AI-driven modelling and forecasting on behalf of the government. It is clear after COVID-19 that the protection of our population will require exceptional technology and collaboration.”

– Dr. Meghan Schaeffer, National Public Health Advisor and Epidemiologist, SAS

Here is the link:

https://itwire.com/business-it-news/data/twelve-ai-predictions-for-2024.html

I find this an interesting list with such a diverse list of suggested futures at all sorts of levels and possible impacts.

It is worth reading through the list to see what you agree with and then to try and see what might have been left out! (Impacts of AI on Law and Accounting professions maybe and what about those who write for a living?)

I see much broader impacts than cited here!

David.

You Have To Love The Delusion The Government Pushes Regarding “Big Tech”

 

I really love the periodic chest thumping we get from our National Government regarding regulation of ‘Big Tech’! 

Federal government moves to shatter big tech’s dominance and ‘anti-competitive’ behaviour

By jared lynch

Technology Editor

5:56PM December 8, 2023

The Albanese government is moving to rein in so-called “big tech” as AI platforms are rapidly adopted across the economy, backing the competition watchdog’s pleas for tighter UK-style laws.

In a potential win for smaller firms, the federal government has agreed “in principle” to the Australian Competition and Consumer Commission’s proposals to stamp out “harmful” behaviour from the world’s top five tech companies.

This includes establishing mandatory codes of conduct designed to prevent anti-competitive conduct, which includes preferencing their own rather than rival app stores and services. The reforms also include strengthening unfair contract terms laws, and forcing digital platforms to shut down scams, fake reviews and harmful apps.

ACCC acting chair Catriona Lowe said the changes would ensure a “fairer and more transparent” system, particularly for small and medium-sized ­businesses.

“Companies like Google, Apple, Meta, Microsoft and Amazon supply services that most Australians use every day. These digital platforms are often unavoidable ‘gatekeepers’ between Australian businesses and consumers, giving them huge influence across the whole economy,” Ms Lowe said.

“By promoting competition in these markets, the reforms will also give consumers more choice, higher quality services, and fairer terms and conditions.”

The ACCC called for the changes as part of its five-year inquiry into the market power of the digital platforms. It has been providing reports every six months on the progress of the inquiry, which began in 2020.

The regulator has highlighted Britain as a potential model in an effort to force Apple and Google to open their app stores to more developers and competitors. Germany, Japan and the EU have also tightened regulations for digital platforms.

While Apple began preparing late last year to allow applications from outside its lucrative App Store to be downloaded onto iPhones and iPads, in response to new EU laws, Ms Lowe said changes in one jurisdiction rarely extended to another.

“It is critical that the Australian government works quickly to implement these reforms so that consumers and small businesses aren’t left behind,” she said.

The new regulatory regime would work alongside Australia’s existing competition laws. Each code would introduce targeted obligations to address the types of anti-competitive conduct most relevant to that service.

Anti-competitive conduct highlighted in the ACCC’s latest report includes self-preferencing, where platforms favour their own services in search results and app stores, and “tying” or forcing developers to use app stores’ in-app payment systems, which charge commissions of up to 30 per cent.

The ACCC also cited “unfair and arbitrary treatment” of smaller businesses that rely on the platforms to reach customers.

ACCC chair Gina Cass-Gottlieb criticised the “ecosystems” of digital platforms late last month, saying they locked in consumers and limited choices, while collecting vast amounts of data. She said the ACCC’s reforms would ensure technology benefited business and consumers.

“Consumers who use multiple products from a single digital platform may be forced to agree to unfavourable terms and conditions and/or accept unpalatable data collection practices due to a lack of suitable alternatives or because it is simply too inconvenient or costly to move out of that ecosystem,” she said.

“Our proposed reforms include a call for targeted consumer protections and service-specific codes to prevent anti-competitive conduct by particular designated digital platforms.”

Meta, which owns Facebook, said: “There is nothing inherent in digital ecosystem businesses that present greater risks of anti-competitive con­duct or consumer harm than in other sectors.”

Microsoft said the ACCC needed to continue its “market-specific analysis when considering firms’ expansion activities and conduct”, while Apple and Amazon said the expansion into smart devices was “relatively recent”. Google said there were consumer advantages from its products working well together.

More here:

https://www.theaustralian.com.au/business/federal-government-moves-to-shatter-big-techs-dominance-and-anticompetitive-behaviour/news-story/2fd675cd7a354645e47982a9b371f446

Surely this is just a game! Without the services provided by ‘big tech’ life would be hardly recognisable. All regulation will be mostly with the agreement of the regulated and the regulators know it!

My view is that ultimately we will get the services we are given where the tech majors have an interest in their provision and where there is some money to be made. The truth is that overall Australians are a profit centre for big tech, one way or another, and that we just have to play along wit the game of providing our data to get the services we want!

Sure we can opt out but it is not clear just how well that actually goes and whether many are prepared to put in the effort to get there. Most seem happy with the deal as it is presently constructed and simply could not be bothered to change!

If you have managed to detach yourself send me some snail mail – while that is still possible- and I will share your approach on the blog!

I reckon our national relationship with ‘big tech’ is almost the definition of symbiosis and I am not expecting change any time soon – despite the huffing and puffing!

David.

I Wonder Just How Effectively Google Is Going To Be Able To Monetize AI?

         This appeared a few days ago:

     Google’s new AI chatbot beats  OpenAI, human experts in tests

John Davidson Columnist

Dec 7, 2023 – 2.00am

Google has launched what it says is the most advanced artificial intelligence invented, Gemini, leapfrogging OpenAI in the race to render human brains redundant.

The “multimodal” AI, trained from the ground up to handle questions asked via a mix of audio, photos, video and text, outperformed every other AI including OpenAI’s GPT-4 model in 30 out of the 32 most popular industry benchmarks, Google officials claimed.

Google’s new AI, Gemini, answers questions asked using a mix of voice, video, photos and typed text. 

It was also the first AI to outperform human experts in a key benchmark known as the Massive Multitask Language Understanding (MMLU) test, Google said.

The most advanced version of the AI, known as Gemini Ultra, will not be available until next year, but an intermediate version known as Gemini Pro will begin to power Google’s free chatbot, Bard, as of Thursday, Google said.

The version of Bard powered by Gemini Ultra, known as Bard Advanced, might well be a paid-for service.

“We’ll explore what monetisation might look like, but we don’t have anything specific on that right now,” Sissie Hsiao, vice president in charge of Assistant and Bard at Google, said.

A third, cut-down version of the AI, Gemini Nano, will appear in Android phones starting with Google’s Pixel 8 Pro phone, to answer complex voice, video, photo and written questions on the phone itself, without the need for an internet connection.

In a video demonstration, Gemini identifies that the photo is homework, marks it, and explains the errors.  

In a global launch event, the company showed off Gemini Ultra performing a range of tasks which, until now, were generally reserved for humans.

In one pre-recorded demonstration, Gemini Ultra was shown a photo of a child’s physics homework, and was able to read it, mark it, and explain the maths and physics errors the child had made, going into levels of detail far beyond what most parents would be capable of.

In another demonstration, two objects were held up in front of a webcam – an orange and a fidget spinner – and the AI was able to identify them both and explain that citrus and the spinner had something in common: they both could be “calming”.

Eli Collins, the vice president in charge of product at Google DeepMind, which developed Gemini, said one of the main features of Gemini was it was less likely to “hallucinate” than other AIs.

“Improving the accuracy of responses was one of the core training objectives of the model. When we talk about getting a better score on these benchmarks, it’s often a result of improving Gemini’s ability to reason and to answer questions factually,” he said.

(And, indeed, the Google search engine does contain plenty of references to the “calming” effects of both citrus and fidget spinners.)

When the orange was replaced by a Rubik’s Cube, the AI identified they were both examples of toys that adults, as well as children, play with.

Coding edge

Google also showed off Gemini figuring out what a complex join-the-dots puzzle was depicting, before anyone even joined the dots (“This is a picture of a crab,” the AI pre-empted.) The AI watched as someone performed a simple sleight of hand with a ball and three cups, and correctly predicted the ball would be in the left cup.

More here:

https://www.afr.com/technology/google-s-new-ai-chatbot-beats-openai-human-experts-in-tests-20231206-p5epgr

I gave Bard a run with the question: “What is the best antibiotic for tonsillitis?”

The core part of the answer given was spot on I reckon:

“I'm not able to provide medical advice, so I can't tell you the best antibiotic for tonsillitis. However, I can share some information that may help you:

Most cases of tonsillitis are caused by viruses, not bacteria. Therefore, antibiotics are not usually effective. However, in some cases, tonsillitis can be caused by bacteria, such as group A streptococcus. In these cases, antibiotics can be helpful.” And so on!!!

Good disclaimer and then reasonable answer!

The question with all this really resolves around how these tools should be best used and how we can sensibly evaluate answers we are given – especially in domains where we are not expert

I found this response interesting: What is the likely outcome of the war between Israel and Hamas?

Answer: The conflict in Israel and Gaza is complex and changing rapidly. If you'd like up-to-date information try using Google Search.

My reaction: “bloody clever response!”

Frankly I reckon we are now in the situation where we can get a reasonable answer to most reasonable questions in less than a minute and I find that amazing,

As a second year med student this would have been really useful to study from!

What is the course of the vagus nerve in female humans?

Answer:

The vagus nerve, also known as the 10th cranial nerve, plays a crucial role in various bodily functions, including digestion, heart rate, and breathing. Its course is similar in both male and female humans, with some minor variations.

Here's a general overview of the vagus nerve's course:

Origins:

• Starts in the brainstem, specifically the medulla oblongata.
• Has two roots: a larger motor root and a smaller sensory root.

Path:

• Exits the skull through the jugular foramen, along with other cranial nerves and blood vessels.
• Descends through the neck, alongside the carotid arteries and jugular veins.
• Enters the thorax and passes through the superior mediastinum.
• Branches out in the thorax and abdomen, innervating various organs.

Branches and innervation:

• Pharyngeal branches: Supply the muscles of the pharynx, responsible for swallowing.
• Laryngeal branches: Control the muscles of the larynx, involved in voice production.
• Cardiac branches: Regulate heart rate and rhythm.
• Pulmonary branches: Innervate the lungs, contributing to breathing and airway reflexes.
• Esophageal branches: Aid in esophageal peristalsis, the movement of food down the esophagus.
• Celiac branches: Supply the stomach, intestines, pancreas, and liver, influencing digestion and nutrient absorption.

Minor variations in females:

• Due to anatomical differences in the chest and abdomen, the specific course of the vagus nerve might show slight variations in females compared to males.
• For example, the vagus nerve may pass closer to the reproductive organs in females.

Additional notes:

• The vagus nerve is a complex structure with numerous branches and functions. This is a simplified overview of its course.
• For a more detailed understanding, it's recommended to consult anatomy textbooks or consult with a healthcare professional.

Just a great answer and I admit as good as one might have hoped for!

I reckon I can now flog all my textbooks – and fast - while they are worth something!  Really books are now obsolete except for reading for pleasure away from screens etc.!

As for how Google makes money with all this, I have no idea!

David.

AusHealthIT Poll Number 726 – Results – 10 December, 2023.

Here are the results of the poll.

Do You Believe The myHealthRecord Has Improved Patient Engagement With, And Understanding Of, Their Health Information?

Yes                                                                                  1 (3%)

No                                                                                  35 (90%)

I Have No Idea                                                               3 (7%)

Total No. Of Votes: 39

A clear outcome with a huge majority feeling that the myHR has added pretty much zero to their engagement with their health information.

Any insights on the poll are welcome, as a comment, as usual!

A good number of votes. But also a very clear outcome! 

3 of 39 who answered the poll admitted to not being sure about the answer to the question!

Again, many, many thanks to all those very few who voted! 

David.