AI in Manufacturing: Scale from Pilot to ROI | Enterprise Playbook
What Is Natural Language Processing? - TechRadar
Natural language refers to the regular speech and text that we use to communicate with each other. Natural Language Processing (NLP) is a branch of artificial intelligence (AI) that enables computers to understand, interpret, and generate human language.
NLP bridges the gap between human communication and computer understanding by combining computational linguistics with machine learning, explains Arturo Buzzalino, Chief Innovation Officer, Epicor.
"AI includes other domains besides NLP, such as computer vision which deals with analysis and generation of images, but advances in NLP in the last few years have been at the heart of the current AI revolution," says Stefan Leichenauer, VP of Engineering, SandboxAQ.
You may likeDescribing NLP as the analysis and generation of natural language with computers, he says, it is the use of Large Language Models (LLMs) and chatbots that are driving a lot of the excitement around the subject.
NLP and LLMsDrilling down further, Volodymyr Kubytskyi, Head of AI in MacPaw, says popular LLMs like OpenAI's ChatGPT or Google's BERT, are trained on massive amounts of text data, allowing them to grasp not just individual words but context, nuance, and even creativity in language.
He argues that it is these LLMs that have pushed NLP to new heights, enabling machines to generate coherent, human-like text, summarise long documents, translate between languages, and even engage in meaningful dialogue. By leveraging these models, NLP can now do things that seemed impossible a few years ago, like writing essays or answering complex customer inquiries in a natural, flowing manner.
"LLMs are the engine that's driving much of today's progress in making machines capable of human-like conversations," says Kubytskyi. "This is AI meeting language at an incredibly sophisticated level."
Why should businesses care about NLP?Leichenauer says because natural language is the way we communicate with each other, a lot of our business operations are encoded in natural language.
"Our reports and presentations, our internal memos and emails, and all of our customer communications are written in natural language," says Leichenauer. "NLP techniques can accelerate and automate workflows involving all of these things."
Building on this Buzzalino explains businesses should care about NLP because it allows them to extract meaningful insights from unstructured text data like customer reviews, emails, and social media posts.
NLP, he says, can help automate tasks such as customer support through chatbots, sentiment analysis for market research, and efficient document processing, thereby improving efficiency and enhancing customer engagement.
Sukh Sohal, Senior Consultant at Affinity Reply agrees. He says NLP brings real impact to businesses by transforming how they engage with customers, handle data, and even communicate internally.
"Imagine an AI that can analyze thousands of customer messages in minutes, picking up on common issues, emotions, or trends," says Sohal. "For companies, NLP can be the difference between overwhelming customer service demands and an efficient, responsive operation."
He says NLP lets businesses automate repetitive tasks, improve customer experience, and respond dynamically to feedback while freeing up human teams for tasks that require real insight.
Kubytskyi is excited about the use of LLMs and how it's elevating these NLP capabilities. For instance, he says, customer service bots
"This level of understanding allows businesses to offer personalized, responsive services without sacrificing efficiency," says Kubytskyi.
NLP applicationsNLP has become so integrated into our lives that we often overlook it.
Buzzalino points to virtual assistants like Siri and Alexa that understand voice commands, customer service chatbots that handle inquiries, machine translation services like Google Translate, sentiment analysis tools that gauge public opinion on social media, and text analytics systems that extract key information from large volumes of documents, as some real-world applications of NLP.
One real-world application of NLP that strikes Leichenauer is as a smart assistant for writing code. This enables developers to operate more efficiently and also allows for low-code and no-code solutions that are more powerful than before.
How does NLP work?Unlike traditional computing, which relies on straightforward commands, NLP involves teaching machines to grasp the subtleties and quirks of human language, including context, tone, and meaning, says Sohal. It's how AI moves from rigid rule-following to more intuitive understanding, opening up new ways for tech to interact with us in a more "human" way.
NLP is built on two key components. There's Natural Language Understanding (NLU), which analyses input to extract meaning and intent, and Natural Language Generation (NLG), which produces responses based on context and system logic, says Dan Balaceanu, Co-Founder & Chief Product Officer at DRUID AI.
For example, when a user requests to "book a flight to London," NLU identifies "book" as the action and "London" as the destination, while NLG generates a follow-up response, like "I found a flight to London for £220. Would you like to book it?"
Technically speaking, Sohal says, NLP works by breaking language down into patterns computers can recognize. It starts with tokenization, where sentences are split into words or smaller chunks. Then, grammar and structure are analyzed to understand the relationships between words.
Semantics come next, where computers use massive data to grasp meanings, even for slang or idioms. Finally, context and intent are added through machine learning, especially deep learning. "Here, NLP models learn from large datasets to identify emotions, requests, or subtleties in language, making responses more human-like," says Sohal.
Balaceanu adds this process standardizes vocabulary by reducing words to their root forms and filtering out common words that add little meaning, which helps to identify the real intent of the prompt that it should respond to, and how it should answer.
He adds that to improve the accuracy of the responses, NLP leans on machine learning techniques, such as deep neural networks, and models like transformers such as BERT.
"For NLP systems to respond accurately, they are trained on vast datasets that include diverse language patterns, grammar rules, and sentence structures, covering a range of possible queries and responses," adds Arunkumar Thirunagalingam, Manager, of Enterprise Data Management at Santander Consumer USA.
He says this training involves machine learning models and deep learning techniques that expose the AI to various linguistic scenarios, enabling it to recognize intent, context, and nuances. Over time, and with continuous learning from large, representative datasets, AI systems become more adept at handling complex language tasks and providing relevant, human-like responses.
Nuance Integrates Generative AI-backed Scribe Into Epic EHRs
Dive Brief:Proponents of generative AI technology in healthcare say it could drastically improve medical care delivery, operations and research. Startups with generative AI solutions in healthcare delivery and life sciences have collectively earned more than $20 billion in funding, according to a recent market analysis by health tech-focused VC firms.
Notetaking is one of the most developed spaces for generative AI in healthcare, as companies like Nuance, Suki and Robin vie to leverage new technologies to solve one of the oldest problems in healthcare — onerous documentation requirements on physicians.
Nuance first introduced DAX Express in March. The scribe combines its own AI with OpenAI's large language model GPT4, the latest iteration of the generative AI technology that backs wildly popular internet chatbot ChatGPT. Large language models are trained to recognize and respond to text based off data they scrape from the web.
DAX Express is notably faster than old Nuance products, turning around notes in seconds instead of hours, the company says.
That's because DAX Express is fully automated, taking out the human reviewer for quality control in Nuance's existing medical scribe service, Dragon Ambient Experience. The automation raises questions about quality, given AI still has a ways to go in addressing some limitations in text generation.
Nuance has yet to publish accuracy measures for its AI-backed transcription tools.
"Nuance focuses on the outcomes that customers achieve. We will publish outcomes data when the product is generally available," a Nuance spokesperson said.
Critics are airing concerns that AI companies move too fast to implement their technology in hospitals without proper oversight. Such fears have given rise to some safety-focused organizations advocating implementation slow down, including a new transatlantic Responsible AI in Healthcare consortium.
Microsoft acquired Nuance in 2021 for almost $20 billion, two years after it first partnered with the voice-to-text company. The deal doubled Microsoft's total addressable market in the healthcare provider space.
Microsoft has been working separately with Epic on generative AI. Microsoft recently unveiled plans to embed generative AI into Epic records, for use cases like automatically drafting replies to patient messages and more easily querying databases.
Microsoft's Nuance Adds GPT-4 AI To Its Medical Note-taking Tool
An AI medical scribe platform is the latest entrant into the AI arms race. On Monday, Microsoft-owned Nuance Communications announced it is integrating GPT-4 into its Dragon Ambient Intelligence platform, which is used by hospitals around the country to ease doctor workloads by using AI to listen to patient-provider conversations and write medical visit notes.
Starting this summer, all providers currently using DAX or Dragon Medical One will be eligible to apply for an early adopter program for DAX Express, which bypasses the human reviewer used as a quality control in the current DAX product, and returns fully AI-generated notes within minutes of a patient visit. This move marks a decisive acceleration of Nuance's timeline for transferring all of the responsibility of drafting notes to AI.
"We're getting much more aggressive," Peter Durlach, Nuance's chief strategy officer, told STAT. Right now, there are less than 100 physicians participating in pilot tests of Nuance's fully AI-powered express mode. Last month, Nuance told STAT that the company had hoped to move into a beta phase by this summer, with a goal of enrolling 400 physicians.
It's not clear what role the GPT-4 technology — which has drawn significant buzz for upgrading the abilities of technologies like ChatGPT — will play in the note-taking model. "All I know is that the team has been very firm in saying there's some magic by combining the two pieces together," Durlach said.
DAX promises to reduce physician burnout by relegating the hours of work doctors spend writing appointment notes to an artificial intelligence platform. The AI listens to the doctor-patient visit, transcribes the conversation, and composes a draft note. In most cases, that draft note is checked by quality assurance staffers before being sent on to the doctor for review. That review process results in a lag of a few hours.
However, the new GPT-4-powered DAX Express is fully automated, eliminating humans from the process until a note gets to a physician for review. The tool provides "an immediate and highly accessible entry point for healthcare organizations to adopt at scale a new generation of AI-powered applications," according to the Nuance press release. The company had been working on transitioning to the fully AI-powered express mode version of DAX for years, starting with its alpha pilot and its planned beta program. Previously, Nuance's stated vision for that "full-service" tool included the option for physicians to bounce an AI-drafted note back to Nuance's reviewers for another set of human eyes.
This option will still exist for users of the express mode of DAX that was being developed through the alpha and beta testing programs. But those who apply to join the newly announced DAX Express with GPT-4 pilot won't have that option for human review.
Paddy Padmanabhan, chief executive of Damo Consulting, a health technology advisory firm in Chicago, said Nuance is seizing an opportunity to jump ahead in the race to build AI voice capabilities for health care providers. Amazon shut down a program late last year to harness its Alexa-branded AI to build voice applications, leaving Nuance to battle with a group of scrappy start-ups that have powerful technologies, but lack its reach and customer base in health care.
"The field is open for Nuance to really accelerate their product development and deployment — and their market share," Padmanabhan said. He added, however, that moving too aggressively could also backfire. "Anything to do with health care has a heightened sensitivity to risk," he said. "The whole point of the technology is to make it as accurate as possible, and minimize corrections that need to be made after the encounter, and that's a work in progress."
Durlach said that Microsoft, and especially Nuance, are being careful of the trust relationship with their clients and are clear-eyed about the powerful new technology. "On one hand, it really is amazing and the GPT-4 makes ChatGPT look like a toy; on the other hand, it makes mistakes and it hallucinates and it omits things," he said. "[So] we're picking applications that we think make sense and where the technology comes to bear in a way that makes sense that if it makes a mistake or two, not going to hurt anybody."
Previously, clinicians had to log hundreds of visits with DAX and progress across DAX's "AI readiness score" to get greenlit for express mode. But Durlach said that any user is eligible to apply to use the new DAX Express, whether they've used DAX before or not. "Part of the reason we accelerated the launch is we believe the vast majority of the DAX users can go right to automated," said Durlach.
However, the accelerated release poses questions about the level of independent scrutiny applied to the product, especially because it is being combined with a powerful but still experimental technology such as GPT-4.
"I would be way more comfortable if [Nuance] worked with their key customers or the early adopters to make their evaluations public," said Nigam Shah, professor of biomedical informatics at Stanford University. Physicians and patients are being asked to trust that Nuance, a multibillion-dollar company racing to out-flank competitors, is doing enough testing to ensure that its product is reliable and beneficial.
Shah said the most useful step would be to test Nuance's product on a publicly available dataset of patient recordings supplied by health systems, so that the assessment of its performance could be truly independent. Since a company is unlikely to pay for such data, Shah said, other parties who want to benefit from the use of the technology would need to supply the data and establish performance standards.
"If we don't do that, we're going to hold ourselves back," he said. "And lots of snake oil will get sold."
This story has been updated to clarify the role of physician review in GPT-4-powered DAX Express.
This story is part of a series examining the use of artificial intelligence in health care and practices for exchanging and analyzing patient data. It is supported with funding from the Gordon and Betty Moore Foundation.
Comments
Post a Comment