Chatbot breakthrough in the 2020s? An ethical reflection on the trend of automated consultations in health care Medicine, Health Care and Philosophy

Eighty-two percent of apps had a specific task for the user to focus on (i.e., entering symptoms). Personalization was defined based on whether the healthbot app as a whole has tailored its content, chatbot in healthcare interface, and functionality to users, including individual user-based or user category-based accommodations. Furthermore, methods of data collection for content personalization were evaluated41.

• Customized chat technology helps patients avoid unnecessary lab tests or expensive treatments.
• There will be a temptation to allow chatbox systems a greater workload than they have proved they deserve.
• Other applications in pandemic support, global health, and education are yet to be fully explored.
• Cancer has become a major health crisis and is the second leading cause of death in the United States [18].
• They can automate bothersome and time-consuming tasks, like appointment scheduling or consultation.

After starting a dialogue, the chatbot extracts personal information (such as name and phone number) and symptoms that cause problems, gathering keywords from the initial interaction. The specific data used for the referral analysis is available in a dedicated GitHub repository. The qualitative feedback data will not be publicly available because the individuals did not provide explicit consent for the public sharing of this feedback data. “What doctors often need is wisdom rather than intelligence, and we are a long way away from a science of artificial wisdom.” Chatbots lack both wisdom and the flexibility to correct their errors and change their decisions. To further cement their findings, the researchers asked the GPT-4 another 60 questions related to ten common medical conditions.

Chatbots in Healthcare: How Hospitals Are Navigating the Pros and Cons

Electronic health records have improved data availability but also increased the complexity of the clinical workflow, contributing to ineffective treatment plans and uninformed management [86]. For example, Mandy is a chatbot that assists health care staff by automating the patient intake process [43]. Using a combination of data-driven natural language processing with knowledge-driven diagnostics, this chatbot interviews the patient, understands their chief complaints, and submits reports to physicians for further analysis [43]. Similarly, Sense.ly (Sense.ly, Inc) acts as a web-based nurse to assist in monitoring appointments, managing patients’ conditions, and suggesting therapies.

This review showed that there is a lack of evidence assessing the effectiveness and safety of chatbots. Further, they should undertake more studies in developing countries and recruit large, clinical samples given the lack of such evidence, as found in the current review. Accordingly, the high risk of bias and low quality of evidence may reduce the validity of the findings and their generalizability. Unfortunately, we cannot draw a definitive conclusion regarding the effect of chatbots due to the high risk of bias in the evidence.

Retrieve Patient Data

Capacity’s conversational AI platform enables graceful human handoffs and intuitive task management via a powerful workflow automation suite, robust developer platform, and flexible database that can be deployed anywhere. Here are the pros and cons of using a chatbot in hospitals or other healthcare facilities. Moreover, training is essential for AI to succeed, which entails the collection of new information as new scenarios arise. However, this may involve the passing on of private data, medical or financial, to the chatbot, which stores it somewhere in the digital world. For all their apparent understanding of how a patient feels, they are machines and cannot show empathy. They also cannot assess how different people prefer to talk, whether seriously or lightly, keeping the same tone for all conversations.

As we balance the allure of AI and the need to protect people’s health, medical chatbots have the potential to improve access to health information—especially when it comes to health issues people typically don’t like to discuss. Chatbots can encourage people to seek help sooner and talk openly about their health. Recent findings demonstrate that ChatGPT is already capable of delivering highly relevant and interpretable responses to medical queries. Medical chatbots can offer fast, remote information to millions of people simultaneously. They may also help streamline healthcare services, reducing some of the current pressures on staff. Considering their capabilities and limitations, check out the selection of easy and complicated tasks for artificial intelligence chatbots in the healthcare industry.

Of the 180 questions asked for GPT-3.5, 71 (39.4%) were completely accurate, and another 33 (18.3%) were nearly accurate. Roughly 8% of questions were completely incorrect, and most answers given an accuracy score of 2.0 or less were given to the most challenging questions. Most responses (53.3%) were comprehensive to the question, whereas only 12.2% were incomplete. The researchers note that accuracy and completeness correlated across difficulty and question type. This story is part of a series on the current progression in Regenerative Medicine. Participants reported that while consultations with doctors were perceived as more accurate, reassuring, trustworthy, and useful, chatbot consultations were considered easier and more convenient.

Chatbots can improve the quality or experience of care by providing efficient, equitable, and personalized medical services. We can think of them as intermediaries between physicians for facilitating the history taking of sensitive and intimate information before consultations. They could also be thought of as decision aids that deliver regular feedback on disease progression and treatment reactions to help clinicians better understand individual conditions. Preventative measures of cancer have become a priority worldwide, as early detection and treatment alone have not been effective in eliminating this disease [22]. Physical, psychological, and behavioral improvements of underserved or vulnerable populations may even be possible through chatbots, as they are so readily accessible through common messaging platforms.

Study Selection

We have yet to find a chatbot that incorporates deep learning to process large and complex data sets at a cellular level. Although not able to directly converse with users, DeepTarget [64] and deepMirGene [65] are capable of performing miRNA and target predictions using expression data with higher accuracy compared with non–deep learning models. With the advent of phenotype–genotype predictions, chatbots for genetic screening would greatly benefit from image recognition. New screening biomarkers are also being discovered at a rapid speed, so continual integration and algorithm training are required.

Case in point, people recently started noticing their conversations with Bard appear in Google’s search results. This means Google started indexing Bard conversations, raising privacy concerns among its users. So, despite the numerous benefits, the chatbot implementation in healthcare comes with inherent risks and challenges.

This is a chat messaging service for health professionals offering assistance with appropriate drug use information during breastfeeding. Promising progress has also been made in using AI for radiotherapy to reduce the workload of radiation staff or identify at-risk patients by collecting outcomes before and after treatment [70]. An ideal chatbot for health care professionals’ use would be able to accurately detect diseases and provide the proper course of recommendations, which are functions currently limited by time and budgetary constraints. Continual algorithm training and updates would be necessary because of the constant improvements in current standards of care. Further refinements and testing for the accuracy of algorithms are required before clinical implementation [71]. This area holds tremendous potential, as an estimated ≥50% of all patients with cancer have used radiotherapy during the course of their treatment.

Agentic AI: Driving Autonomy In Asset-Intensive Industries

Official government figures are less dire, with the Bureau of Labor Statistics reporting that the median tenure of a manufacturing worker dropped from 5.9 years in 2014 to 4.9 years a decade later. But regardless, manufacturing workers are coming to work each day with a lot less expertise than those who worked similar jobs decades ago. As the AI hype cycle cools and business pressure mounts, now is the time to become practical.

In other words, organisations will likely be at least two years behind the competition, at risk of losing money on unnecessary asset replacements, and possibly suffering more frequent unplanned downtime and wasted energy. All this while falling further behind on productivity and sustainability – not something many businesses want to be seen doing in 2025. As AI solutions gain traction in manufacturing, they signal a broader movement—one in which intelligent tools amplify human expertise, making specialized knowledge more accessible and operations more efficient. From automating quality control and predicting equipment failure to optimizing supply chains and enabling intelligent surface preparation through these tools, AI is ushering in a new era of precision, efficiency, and adaptability. As manufacturers adopt AI-driven systems to optimize operations and reduce downtime, these technologies are becoming central to the evolution of Industry 4.0, where data and automation reshape how products are made and maintained. This guide by 10X Engineered Materials covers use cases, benefits and the future of AI in manufacturing.

How AI is improving quality control and reducing defects

The result is more uniform surface finishes and minimized abrasive use—two critical factors for quality and cost efficiency in aerospace, automotive, and heavy equipment maintenance industries. In abrasive blasting operations, where components like nozzles, hoses, and compressors degrade with use, predictive models can anticipate when performance thresholds are nearing critical limits. This knowledge can reduce system downtime by 35%, lower maintenance costs by 28%, and provide an additional layer of worker safety.

Google launches TPU monitoring library to boost AI infrastructure efficiency

• By integrating machine learning algorithms and vision systems into operations, companies can unlock new levels of precision and efficiency previously unattainable.
• Danone has a partnership with the Singapore Economic Development Board (EDB), with a joint initiative around ‘digital health’ since 2018.
• But regardless, manufacturing workers are coming to work each day with a lot less expertise than those who worked similar jobs decades ago.
• IT leaders should update their data and AI governance frameworks to cover field operations use cases that require sharing sensitive data with third-party LLMs and AI agents.
• As AI integrates into machines and control panels, workers will learn to operate and troubleshoot systems using new interfaces like augmented reality, natural learning commands, and AI-driven human-machine interfaces.
• Even when SaaS platforms announce agentic experiences, data teams should evaluate whether data volume and quality on the platform are sufficient to support the AI models.

Using large language models and machine learning, GenAI platforms moved beyond the responses of chatbots to evolve and grow (“learn”) from their interactions with people. However, despite the apparent benefits, many companies still resist adopting these “no-brainer” solutions. Instead, they continue using spreadsheets that merely “tick the box” or they replace assets on a fixed schedule or after breakdowns, when IoT based tracking would be more effective. This is costing them dearly – in downtime, energy waste, and missed opportunities for improvement. A year ago, Danone signed with US software giant Microsoft to “explore” a multi-year collaboration to integrate AI throughout its operations. Ask about how AI can provide value, then learn how to present the concepts without drumming up fears of job loss.

For example, take a service provider like a facility management or catering company managing critical assets such as refrigerators and freezers in hospitals or catering environments. Additional review is needed when there are data privacy and regulatory concerns. IT leaders should update their data and AI governance frameworks to cover field operations use cases that require sharing sensitive data with third-party LLMs and AI agents. Some field operations perform a set of common tasks at different locations and must adapt to local conditions and requirements. For field operations that perform a wider variety of work types in highly differentiated conditions, AI agentic experiences partner with field engineers to provide real-time information and guidance.

Agentic AI: Your Personal Assistant

• And there’s no doubt that the organisations that will truly thrive in the coming years are not the ones waiting for perfect conditions or more case studies.
• Before the generative AI boom kicked off in late 2022, Siemens had more than 1,500 AI experts on staff.
• Managing these assets and ensuring their functionality or industry compliance is still, in many cases, tracked using manual processes – relying on spreadsheets, which are time-consuming and prone to error.
• This level of understanding allows them to identify subtle process shifts and nonlinear relationships that SPC might overlook.

Focus on projects with clear, tangible metrics and ROI to assess the benefits. As with all AI projects, ensuring the system has plentiful, clean data is necessary. Governance and standards must be in place to ensure boundaries around AI use.

Consider a high-throughput line where abrasive blasting is one of several value-added steps. AI can optimize upstream and downstream processes to synchronize with blasting throughput, ensuring balanced workloads and minimal idle time. It can also correlate variables such as abrasive type, pressure, and humidity to final finish quality, recommending adjustments to maximize yield. Instead, it extends their abilities through autonomous, guided actions.

Oracle launches MCP server to power context-aware AI agents for enterprise data

AI-based systems go further by continuously learning from live process inputs and adapting thresholds dynamically. This level of understanding allows them to identify subtle process shifts and nonlinear relationships that SPC might overlook. Artificial intelligence (AI) is no longer a future-facing concept in manufacturing. It is a present-day force accelerating productivity, precision, and decision-making across the industrial sector.

Alaska Airlines lifts ground stop after IT outage affects operations

Before the influence of generative AI, the company had an annual productivity gain target of 3% to 4% across the business. But Koerte says with AI, the hope is to double that aspirational range. Across the broader workforce, there are 460 distinct AI use cases in production today, a figure that excludes the unique AI chatbots that workers can create to perform tasks through a secure system that’s called SiemensGPT.

Ahrefs Adds AI Content Scores to Site Explorer

For instance, Daedalean is using vision-based systems to interpret runway incursions and weather patterns in real time. Similarly, MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) has also developed Air‑Guardian, an AI-enabled copilot that monitors a pilot’s gaze and intervenes when their attention is lacking. In tests, the system “reduced the risk level of flights and increased the success rate of navigating to target points”—demonstrating how AI copilots can enhance safety by assisting with real-time decision-making. As these technologies evolve, they hold the potential to reduce the number of required pilots on aircraft with more than nine passengers from two to one, in alignment with current FAA regulatory pathways. Infosys is transforming ERP deployments by embedding AI across functions like finance and supply chain, enabling faster, more agile rollouts compared to traditional ERP projects. Our AI-powered suite accelerates SAP S/4HANA Cloud transformations by automating complex tasks such as code migration and invoice processing, reducing manual effort and errors.

Rise of the Machines: Inside Hollywood’s AI Civil War

Whether optimizing weather and traffic avoidance systems or refining aerodynamics and structural integrity, AI is set to revolutionize the aviation industry, unlocking a new era of innovation and performance. AI is poised to deliver over $40 billion in value industry-wide by 2033, with a compound annual growth rate (CAGR) of 38.1% between 2024 and 2033. While AI’s applications in aviation are not yet widely scaled, they represent exciting opportunities to enhance safety, reduce costs and improve decision-making across the industry. Today, AI is transforming pilot training by analyzing vast data from modern aircrafts to provide personalized feedback and recommendations. Yet OpenAI reps have hung around film festivals and made entrées to filmmakers and studios, while Sora, the company’s video tool, has been used for a host of films from interesting artists. One of them is Air Head, a micro-short from the Toronto pop collective shy kids.

The New Storage Architecture

Similarly, Airline Pilot Club’s “Amelia” uses AI to tailor training programs to individual pilots. I expect innovations like these will create a shift toward safety and reduce time and costs for pilot training in the industry, especially with a crucial pilot shortage nowadays. Paul Thomas Anderson and Wes Anderson will always be Paul Thomas Anderson and Wes Anderson, of course, and it’s highly unlikely their brand of bespoke film will change. But the AI age might not produce much of a new Andersonian generation if there is no studio ecosystem (or commercial market) to support that type of human-guided film. Given its lower costs, automated media doesn’t need to match human-led work — it need only be good enough.

• Ahrefs says it blends statistical token analysis with neural pattern recognition and can identify specific models such as GPT-4o and Llama.
• The company, whose credits include the Oscar-winning “Parasite,” demonstrated two key AI systems currently integrated into its production workflows.
• Tung emphasized that while AI can support early stages such as ideation and research, its application is limited when it comes to core artistic contributions.
• There’s a radical thought that AI cinema will help the film world conjure not just scenes but people — “digital humans” that will look and move like real actors without any of the pesky concerns of a bad day, or residuals.
• “I co-write with Gemini (or ChatGPT) asking it for ideas, picking the best ones, and shaping them into a simple script,” he wrote.

Get those pieces right, and the black-box ref can fade into a trusted safety net, visible only when it matters most. Outside the United States, soccer offers the most mature example of end-to-end AI officiating. FIFA’s Semi-Automated Offside Technology (SAOT) marries a 500 Hz sensor inside the match ball to twelve roof-mounted cameras that track every limb 29 times per player. A machine-learning model reconstructs a 3-D field and alerts the video assistant referee when an attacker strays beyond the last defender, cutting typical reviews from minutes to roughly 25 seconds. Independent testing places its average error at roughly 3.6 millimetres (just over 1/10th of an inch), small enough to retire most line judges. A 2024 field study of Grand Slam matches also linked the technology to an 8% drop in overall umpire mistakes.

The Google cyberattack prevention tool has been deployed to safeguard widely used open-source tools that power a huge chunk of the internet. You don’t have to scroll for long to come across techno-utopians declaring with no evidence that AGI is right around the corner and that it will cure cancer and save our society. Others are interested in the technology but are concerned that it’s being brought into the industry too quickly, with insufficient planning and protections. “How fast do we roll out AI technologies without really having an understanding of them?” he asked in an interview with Production Designers Collective. “There’s a potential for AI to accelerate beyond what we might be comfortable with, so I do have some trepidation and am maybe not gung-ho about all aspects of it.” While in LA for the event, I visited separately with two of my oldest friends.

It is, in one sense, the ideal illusion, tech creating the next jaw-dropper to induce running or fainting without the requirement of any real-world duct tape to make it happen; you simply snap your fingers and it appears. Data storage in a future world of 5G, AI, and IoT will resemble intelligent, decentralized architectures that provide flexible, scalable solutions, serving unprecedented growth and change in real-time, both today and for many years to come. Innovations such as quantum-safe encryption, DNA storage, and blockchain-enabled content authentication are going to play an increasingly important role in securing, maintaining the integrity of, and scaling storage architectures.

Likewise, teams in finance, travel, or news can benchmark their own AI adoption against these fully automated examples and set internal thresholds with real numbers instead of guesswork. Because the detector lives inside Site Explorer, we hope that users could filter pages by AI share in the future, isolate subfolders and export the numbers. That workflow matters for competitive audits, risk reviews before link outreach and internal quality checks. The process began with Kalshi giving Accetturo a list of themes, ideas and bits of dialogue, Such says. He then turned those concepts into a script and AI prompts — the instructions that tell AI systems what kind of content to generate.

Expanding connectivity increases security risks, demanding zero-trust models, encryption, and AI-driven defenses. Compliance and scalable, sustainable infrastructure are essential for managing data growth and effectively meeting regulatory and privacy challenges. 5G networks enable network slicing, allowing enterprises to create virtual networks tailored to specific workloads or business units. This ensures efficient data handling and supports diverse storage needs across departments and applications. On any given weekend in 2025 a sports fan might zap between three broadcasts and watch artificial intelligence settle pivotal plays in real time.