Author: Professor Danielle McCarthy (PhD, RNutr)
I decided to write this piece because I feel we need to increase our understanding of this domain at pace and I believe the ideal way is to do this is to openly share our experiences.
To ensure we have the confidence to ask questions, make suggestions, be bold with our ambitions, and firm with the rigour with which we work with artificial intelligence (AI), I have composed this personal essay reflecting on the potential role of ChatGPT in the life of a professional nutritionist.
There are three things that I know influence my perspective on this subject which I feel I should be clear on upfront.
- The generative artificial intelligence (gen AI) explored in this reflective piece is ChatGPT 3.5 which is a specific type of gen AI, based on an open system and a large language model. Generative AI is a domain of AI that takes various inputs (e.g. from video, text, images, audio) and generates new content. For example, in recipe development, it could be used to generate a new recipe based on an individual’s stated preferences and specific ingredients.
- I come to this reflective piece with bias from my experience as Chief Health Officer at Spoon Guru, an AI-based nutrition intelligence platform which is based on a closed system and a small language model (ie. it is AI-based, but different to ChatGPT). I believe well designed AI-based tools, which have expert nutritionists involved in both their design and deployment, hold huge potential in supporting our profession to deliver meaningful health impacts at scale both now and in the future. I had a short article on AI and its role in delivering nutritional advice published last year which discusses this specific use case further.
- In this reflection I consider myself in the round – as a human who has two jobs focused on nutrition (I am also co-founder of Nutrition Talent), and as a mum to 3 kids and someone who values having time alone and with friends.
The idea to write and share this reflective piece came after I attended Aliya Porter’s (RNutr) workshop entitled “AI in Nutrition Communications”. In a 2-hour session Aliya took attendees through an introduction to gen AI and guided examples of using ChatGPT in ways that could be relevant to professional nutritionists. When I saw Aliya was delivering this training, I jumped at the chance to attend. I was keen to meet fellow nutritionists interested in the space, and have the opportunity to discuss experiences, concerns and opportunities together.
How might Nutritionists use Generative AI?
- Meal Plans
- Written resources
- Improve reach via search engine optimisation
- Answering emails
- Diary management
- Composing marketing posts
- Personalised lunch suggestions from my fridge (Meal Reveal)
- To access summaries, get up to speed quickly on new topics
- Trends, insights
This list is non exhaustive. It shows several types of new information that can be generated. But are these all useful? What information are they based upon? Are they all suitable for a Nutritionist to use in a professional context? Would we agree with the answers that are generated? What are the unintended consequences of using this technology in these ways? What if the wrong information is being generated? What will using this technology replace? We must collectively consider all these questions (and many more) to retain and/or improve the high-quality standards of our practices today. Frameworks to guide such practices would be helpful, and extensions to codes of conduct may be useful if we are to utilise such technology thoughtfully, safely and effectively.
How can those in the nutrition profession be an active part of the technology transition that is happening at pace?
I believe the phrase often coined in business is “lean in”. We are already in. Technology, devices, digital media, algorithms and wearables are already infiltrating our personal lives. Our professional lives are no different. I believe our profession, and all of those in our profession, not just those working in the tech sector or interested in tech, have experience and insights that could ensure this tech-driven landscape is designed and deployed in ways that assist us in overcoming challenges we face as nutritionists, such as;
My clients are confused and overwhelmed by health messaging, much of which is misleading.
My job these days is more in front of excel than people. I chose this career for food and people, not data crunching.
I am so busy firefighting and reacting to organisational needs that I have no time for the strategic work that needs to be done to make a real difference.
I’m feeling increasingly distant from my fellow professionals as I often work alone.
These are just a few of the challenges we repeatedly hear in our career conversations with our Nutrition Talent community. Approaching AI with an open mind, giving time to learn more and consider how it may help resolve some of these challenges will ensure our (nutrition) domain expertise and applied experiences inform the design and deployment of tech-based solutions. We have a responsibility to ensure we contribute our knowledge to inform and be part of the changes which are happening right now.
What risk does Generative AI pose to our profession?
Thoughts I have heard expressed include;
AI is scary, I don’t trust it. It gives all the wrong information.
Is it going to take our jobs and make nutrition even more confusing to the public?
I don’t understand it, I’m sure it will come and we’ll all be using it in a few years.
From my perspective, the real risk is that fear could hold our profession back from stepping forward and driving responsible AI design and utilisation. We may not realise the full benefit if we do not actively take time to imagine what is possible. Many of us do not feel comfortable or have the time to stick our heads above the parapet on social media, to correct every inaccurate piece of information we see. I would flip reverse the statements above into a lever to encourage us all to face into this change and act. As Aliya demonstrated in the course, there are ways today that ChatGPT could be used to help us reach a wider audience e.g. SEO optimisation. We can apply our critical analysis skills, quality standards to ensure the outputs of ChatGPT are used responsibly. Indeed Aliya shared some new jobs that will arise, if not already in existence, that Nutritionists would have the rights skills for e.g. GenAI reviewers used to fact check all outputs to be utilised by a company, or Prompt engineers, to provide deep considerations into Generative AI inputs to help ensure quality outputs.
How do we proactively manage Generative AI usage within our profession?
My approach here is not to throw the baby out with the bath water and dismiss this area entirely based on negative headlines or fear of the unknown. I fully admit that on a daily basis between work, family commitments and making time for the things I enjoy, I am living and working in a way that is beyond human scale, a term I heard recently in a great podcast by Brené Brown & Esther Perel. So how can GenAI help me take advantage of all that my personal and professional life has to offer?
I actively put effort into not letting fear hold me back. “If I had courage today, what would I do next?” Therefore, my personal reflection is:
If there is potential for this technology to drive efficiency in my world, professionally and personally, in ways that enable me to experience and/or deliver the same or better quality of life and work, rather than shut down the concept, I am open to explore it.
I believe nutritionists are extremely well placed to do this. We have strong critical evaluation skills, experience applying nutrition knowledge in multiple contexts, working with a multitude of different functional teams and adapting to changing environments and evolving science. I believe we should have confidence to contribute our expertise proactively. This will mean experimenting with tools, feeding back, ideating together. If we step back and wait for it to be developed around us, our expertise will not be integral to solutions, the horse will have bolted. This area is changing at speed. We cannot wait for someone else to consider and fix things. If these tools are to be designed and delivered in ways to support, empower and enable our profession and those we service, we must step in, step up and contribute.
Extending existing bias such as gender and cultural inequities needs to be actively tackled. Stigma, such as weight stigma, cannot be extended into these developing propositions. As Nutritionists we need to work actively to ensure this is addressed.
Standards are needed. These may vary by the context in which you are working as a nutritionist. It may vary by sector or area of application e.g. digital marketing vs policy making. Just like the publication guidelines a journal defines for authors, or how regulators publish conditions of use for health claims, so too could clear frameworks help ensure safe use of Gen AI in different contexts. Given the changes are happening at speed, the way in which these are developed would need to match this speed or they will be too little too late.
It’s time to experiment, work together, share findings, make recommendations
I did not learn nutrition science alone. I have never applied my nutrition knowledge alone. I’m not going to start working alone now. But a fear I have is that ChatGPT could generate personal and professional disconnection. Consider for a moment, you have a deadline to create something, a presentation, a strategy, a report. You have a block. Your mind is blank. You’re having trouble getting started. You used to go to make a coffee, go to the water cooler and chat it over with a colleague. Now maybe it’s going on a zoom call for a conversation which sparks your thinking and off you go. What will be next? And let’s be honest, for many, what is already the new default? Ask ChatGPT? Get some ideas from its output first? What does this mean for human connection? Collective ideation? Relationship building based on sharing vulnerabilities? Personal and professional conversations that happen at times of stress, frustration, burning ambition, a shared passion to solve something that’s really difficult ignite new ideas, they build deep empathy and understanding and strengthen relationships. If we just ask ChatGPT, look at what we lose. We need to get conscious with when and how we use such tools, and what this replaces.
Let’s continue to create ideas together, compare notes, have dialogue. Let’s do this alongside new genAI tools and the “botputs” they give us.
The last thing I think we should do as a profession is attach shame to the use of Gen AI. That somehow utilising technology makes you less of a professional. We must remain open and critically engaged with its utilisation. At a time when many feel isolated, and comparative culture is rife, alongside high incidence of imposter syndrome, burnout and busyness, utilising ChatGPT could be very useful. I believe its transparent use, critical evaluation, guidance, frameworks and agreed standards for practice will ensure we are best placed as nutritionists to use these new technologies effectively within our robust practices.
7 Considerations for Nutrition Professionals utilising GenAI
- Consciously address bias: The outputs, as a result of your prompts and the data accessed, could contribute to the bias that already exists. We should actively address this when utilising these tools and share best practice.
- Consider speed and substance: We must consider the unintended consequences of the shortcuts this technology can deliver. What skills do we need to keep strong, what muscles must we keep working as skilled nutritionists? Critical thinking, attention to detail, thorough evaluation, in-depth dialogue.
- Consider ethics: We must apply the same ethical considerations in this domain as any other in which we work e.g. privacy, inclusivity, fairness, robustness, transparency.
- Information trail: Where is the information that is being generated coming from? Are those sources reputable? Are those individuals getting the credit they deserve? The pay they deserve? If utilising ChatGPT in your practice, is this being declared openly as current sources of information are shared?
- Share vulnerabilities, the unknowns: We need to be open with each other. We need to ensure we all have the opportunity to build our understanding and identify best practices in these new landscapes. How do we make sure we learn from each other? The answers ChatGPT generates can sound highly affirmative. Imposter Syndrome, we know from our Nutrition Talent experience, is rife across all sectors in which we work. We must work to support each other in these changing times.
- Human to human connection: We must actively protect this. What happens to those conversations in which you seek help? When you have a block? When you need to download, when you need a deeper understanding? We need to maintain the importance of solving problems together, alongside tech-based solutions.
- Nutritionists involved: We must do everything we can to ensure those with nutrition expertise are involved in the design and deployment of AI-based tools which are designed to impact food and health behaviours.
Final remarks
Let’s not work in a professional vacuum and be siloed by technological developments which could help and/or hinder us in our roles. Let’s explore these openly together, set standards and feel empowered to make decisions on what is right to use, when.
It is now that principles, ethics and ways of working should be being established.
I would like to put a call out to all fellow nutritionists that if this is a topic that you are interested in, connect with me directly (info@nutritiontalent.com) as this is an area we are very passionate about and keen to support our wonderful profession on.
Author: Professor Danielle McCarthy (PhD, RNutr)
Contact us if you have nutrition consultancy or recruitment requirements.
Read next..
Part 2 – Personal experiment #1: Danielle instructs ChatGPT “Can you write me a blog 600 words on considerations for the use of generative AI by nutrition professionals”
Part 3 – Personal experiment #2: Based on Danielle’s original article above, ChatGPT is instructed “Prioritise SEO optimisation in this text without compromising readability. Identify and seamlessly integrate relevant keywords or phrases to enhance online visibility to nutritionists and key stakeholders in the food system”
You’ll get a chance to compare human generated output to GenAI ‘botput’.
Will one reign supreme? Or better together? To be continued….