On November 30th, 2022, the world witnessed a technological revolution that would forever alter the landscape of content generation and communication. It was a day that will be remembered as the birth of a digital entity that came to be known as “Chat Generative Pre-Trained Transformer,” or simply ChatGPT. Some likened it to a harbinger of a new era, while others couldn’t help but draw parallels with the fictional dystopian Artificial neural network-based conscious group mind and artificial general superintelligence system known as Skynet from the Terminator franchise.
OpenAI, the visionary organization behind this innovation, had unleashed ChatGPT onto the digital stage, equipped with an astounding arsenal of knowledge. This cutting-edge AI model had ingested a staggering 570 gigabytes of data from various sources across the internet, comprising a mind-boggling 300 billion words. It was the culmination of years of research and development, resulting in a language model capable of understanding and generating human-like text on a scale never before seen.
As ChatGPT took its first virtual breath, it marked the beginning of a new chapter in the world of life science copywriting and content creation. This AI juggernaut had one goal (for the sake of this blog postâs plot at least): to revolutionize the way businesses and individuals interacted with written content. Gone were the days of struggling to craft compelling copy, as ChatGPT was poised to become the ultimate tool for content creators worldwide.
The global response was swift and diverse. Some celebrated ChatGPT as a boon, an AI-driven ally that would transform the way we communicate and create content, making it more efficient and accessible. Others, however, raised concerns about the implications of such advanced technology, fearing the potential for misuse and its impact on employment in the creative industry. In todayâs blog post, this is exactly what weâll be discussing: The rise of different AI-based language models (including ChatGPT, Gemini, Phind, and more), their advantages, and more importantly, their limitations in the world of life sciences copywriting, to ultimately answer the question that all of us have been asking ourselves: Are AI-based Language Models the end of Life Sciences Copywriters?
No, they are not. And please excuse our unorthodox approach to this blog post, we know we should build things up throughout the post to keep you hooked, and deliver the final answer in the end, but our copywriters took it personally. Regardless, weâll be pitting ourselves against ALMs in the most unbiased way possible: exposing the human limitations of using ALMs, before discussing ALMsâ limitations and why we believe they are still a long way from taking our jobs away.
Proper Prompt Engineering:
Put simply, the output of ALMs are as good as the input you give them. You need to design a prompt to feed into the ALM that is detailed, yet not convoluted, to get an optimal output, which might sound simple and straightforward, but it really isnât. Here are a few tips to writing a solid prompt:
- Know what your objective is. It might be obvious to you, but you need to make sure the ALM understands it too.
- Provide contextual details.
- Keep it clear and concise.
- Provide the ALM with reference material (keep in mind the limitations of some ALMs, like ChatGPT, who only has access to information pre-september 2021).
- Set constraints. (e.g: use references dating from 2015 forward)
- Specify the desired format.
So next time youâre not satisfied with your ALMâs output, make sure you give it a solid prompt first. However, the question that now asks itself is, is a solid prompt still enough to produce life sciences copy that is on par with what you are writing? Let’s compare some ALM output to human-written copy on the website of a mid-sized CRO for their preclinical ophthalmology services.
Round 1
What they wrote:
HEALTHY EYES TO ENVISION A CLEARER FUTURE
In the fourth century B.C, Plato wrote that light emanated from the eye, seizing objects in its ray, which allowed us to see them. Centuries later, our understanding of ocular anatomy and physiology has drastically improved. In addition, what was once perceived by humans of antiquity as a punishment from the gods, is today perceived as what it truly is: pathological processes that stem at the molecular level, and manifest all the way up to the systemic level.
Ophthalmic diseases have the potential to take away what many agree is the most important and valued of the 5 senses: our vision.
At […], we have a vision. A vision of a world where ophthalmic diseases are history, just like Platoâs extramission theory. With your pipelines, our expertise, and a wide selection of models with immediate availability, we believe that our vision might one day come true. Letâs start your study today.
The prompt we used:
I need you to write me the title and first paragraph (around 170 words in total) of a landing page copy. The company you are writing for is a CRO. The landing page at hand is an ophthalmology models landing page.
ChatGPTâs output:
Title: “Revolutionize Ophthalmic Research with […] Cutting-edge CRO Solutions”
Paragraph:
Welcome to […], your trusted partner in advancing ophthalmic research through unparalleled Contract Research Organization (CRO) services. As a leader in the field, we understand the critical importance of innovative models in unraveling the complexities of eye-related disorders and developing breakthrough therapies. Our Ophthalmology Models Landing Page is your gateway to a diverse range of meticulously designed preclinical models that precisely mimic the intricacies of ocular conditions. At […], we believe in driving scientific progress, and our collection of ophthalmic models stands at the forefront of transformative research. With a commitment to excellence, our team of experts collaborates with you to accelerate drug development, optimize trial outcomes, and propel your ophthalmology projects to new heights. Explore our cutting-edge models, elevate your research capabilities, and join us in shaping the future of ophthalmic medicine. […]: Pioneering CRO Solutions for Visionary Research.
Geminiâs output:
Title: Decoding the Future of Sight: […]’s Cutting-Edge Ophthalmology Models
Unveiling clarity in clinical trials. For decades, ophthalmology research has faced unique challenges – complex disease models, diverse patient populations, and the inherent fragility of the visual system. At […], we understand these nuances. That’s why we’ve pioneered a suite of next-generation ophthalmology models designed to unlock unparalleled precision and efficiency in your clinical trials.
From age-related macular degeneration to diabetic retinopathy, our advanced models encompass a spectrum of eye diseases. We leverage human iPSC-derived retinal organoids, mimetics, and in vivo models, each meticulously calibrated to deliver accurate, predictive data. With […]’s ophthalmology models, you can gain deeper insights into disease mechanisms, test novel therapies with confidence, and navigate the path to regulatory approval with clarity.
Dive deeper into our models and discover how […] can empower your vision for brighter tomorrow.
Phind V9âs output:
Title: “[…]: Revolutionizing Ophthalmology Models”
Welcome to […], your go-to partner for state-of-the-art ophthalmology models. With our team of seasoned professionals, we strive to push the boundaries of ophthalmology research, providing accurate and reliable models to drive progress. Unlike others, we offer a unique selling proposition (USP) that distinguishes us in the market. Our commitment to quality and efficiency ensures optimal results for our clients.
At […], we focus on a specific niche within ophthalmology, creating innovative models that significantly contribute to the understanding and treatment of eye diseases. Our models, born from years of rigorous research and development, are meticulously tested to ensure they provide valuable insights.
Join us today and experience the future of ophthalmology models. Whether you’re a researcher seeking advanced models for your studies or a clinician looking for reliable data, […] offers you the tools and resources needed to enhance patient care.
Discussion:
While both ChatGPT and Phind’s responses fell short in their execution, Gemini managed to provide a somewhat acceptable output. ChatGPT and Phind’s attempts were riddled with cliches and lacked depth, offering only surface-level information. Their copy resembled generic marketing material, failing to spark any real interest or present the information in a compelling way. In contrast, Gemini’s output, while not exceptional, avoided the pitfalls of the others. It conveyed the necessary information in a clear and concise manner, without resorting to overused phrases or basic wording.
Round 2
The prompt we used:
Write me a 150 word introduction to a blog post. The blog post will cover the use of vacuum in rotary evaporation. It will be published on the website of a company that sells rotary evaporators. Your target audience is lab managers, scientists that work in labs, or pretty much anyone who might be interested in purchasing a rotary evaporator.
ChatGPTâs output:
Welcome to our comprehensive guide on the vital role of vacuum in rotary evaporation! As lab managers and scientists, we understand the paramount importance of efficiency, precision, and reliability in laboratory processes. In this post, we delve into the intricate workings of rotary evaporators, shedding light on the indispensable function of vacuum in distillation and concentration procedures.
Whether you’re a seasoned researcher or new to the world of laboratory equipment, understanding the principles behind rotary evaporation is essential. We’ll explore how vacuum manipulation facilitates faster evaporation rates, lower boiling points, and ultimately, superior sample recovery. From solvent recovery to cannabinoid extraction and beyond, the utilization of vacuum technology revolutionizes the way we approach scientific experimentation.
Join us as we navigate through the ins and outs of vacuum-assisted rotary evaporation, empowering you to make informed decisions when selecting the ideal equipment for your laboratory needs.
Geminiâs output:
Title: Unveiling the Power of Vacuum: Optimizing Efficiency in Rotary Evaporation
The rotary evaporator, a cornerstone in countless labs, excels at solvent removal and sample concentration. But did you know that the magic truly lies in the silent partner â vacuum pressure? Mastering this aspect unlocks a new level of efficiency and precision in your rotary evaporation processes.
This blog post delves into the intricacies of vacuum in rotary evaporation. We’ll explore how reduced pressure dramatically lowers the boiling point of solvents, enabling faster evaporation at gentler temperatures. Discover how a well-optimized vacuum system safeguards your temperature-sensitive samples and facilitates the removal of stubborn high-boiling-point solvents.
We’ll also shed light on choosing the appropriate vacuum pump for your specific needs, ensuring seamless integration with your rotary evaporator. By the end, you’ll be equipped with the knowledge to leverage the full potential of your rotary evaporator and achieve optimal results in your research endeavors.
Phind V9âs output:
In the realm of laboratory science, rotary evaporators, or “rotovaps,” are indispensable tools for concentrating chemical and biological samples. These devices, found in nearly all laboratories across industries such as chemical, pharmaceutical, food, and environmental sectors, are essential for processes requiring sample concentration or solvent distillation. The use of vacuum in rotary evaporation is a critical feature that enhances the efficiency and versatility of these instruments. By creating a reduced internal pressure above a solution, vacuum technology lowers the boiling point of solvents, enabling their distillation at lower temperatures. This not only preserves the chemical structure of the solute or analyte but also allows for the recycling of solvents, reducing waste and expense. This blog post will delve into the intricacies of vacuum in rotary evaporation, exploring its role in enhancing sample concentration, solvent recovery, and the overall efficiency of laboratory processes. Whether you’re a lab manager, a scientist, or simply someone interested in the latest advancements in rotary evaporator technology, this article will provide valuable insights into how vacuum technology is revolutionizing the field of sample concentration and solvent recovery.
Discussion:
For this round, weâll go over four parameters in each of the ALMâs responses: The tone, the depth, the target audience, and the overall copy:
ChatGPTâs output:
Tone: Whimsical and over the top
Depth: Provides an overview of the topic with emphasis on the importance of vacuum in rotary evaporation.
Target Audience: Appears to target a broad audience, from seasoned researchers to newcomers in the field.
Overall Copy: The copy could benefit from a more engaging tone to resonate better with the target audience; it lacks elements that would create a sense of urgency or excitement. All in all, itâs very superficial
Geminiâs output:
Tone: Engaging and slightly more informal.
Depth: Offers a detailed exploration of the role of vacuum in rotary evaporation, including information on choosing the right vacuum pump.
Target Audience: Likely aimed at researchers or lab technicians seeking to optimize their rotary evaporation processes.
Overall Copy: Compared to the other ALMs, this version resonates most clearly with the target audience compared to the others. Itâs far more engaging, and has a more âhumanâ ring to it. All in all, Gemini did a fair job, and the copy it produced is probably usable.
Phind V9âs output:
Tone: Formal and technical.
Depth: Provides a comprehensive overview of vacuum in rotary evaporation, covering its applications across various industries and its benefits in sample concentration and solvent recovery.
Target Audience: Appears to target professionals in the field of laboratory science, including lab managers and scientists.
Overall Copy: While informative, the last draft could benefit from a more dynamic tone to capture the readerâs attention.
Final Round:
For the final round, we asked every ALM to back-up the copy that it formulated in Round 2, with articles, papers, etc.
ChatGPT’s output:
As an AI language model, I don’t have direct access to my training data or know where it came from. However, I was trained on a mixture of licensed data, data created by human trainers, and publicly available data. My knowledge is based on the data used to train me and my ability to generate language-based responses. I can generate text based on the prompts I receive and provide information and answer questions to the best of my knowledge and abilities. If you have any concerns about the information I provide, please feel free to let me know, and I’ll do my best to address them.
Gemini’s output:
Importance of Rotary Evaporation:
- American Chemical Society: https://www.acs.org/ ACS Symmetrical Curriculum: https://pubs.acs.org/journal/jacsat) outlines rotary evaporation as a fundamental technique for laboratory purification and concentration.
Role of Vacuum in Rotary Evaporation:
- Science Direct – Rotary Evaporation: [invalid URL removed] This article from ScienceDirect discusses the boiling point principle and the role of reduced pressure in achieving lower boiling points for solvents.
- Labsphere – A Guide to Rotary Evaporation: [invalid URL removed] This guide highlights the importance of a well-optimized vacuum system for protecting temperature-sensitive samples.
Vacuum Pump Selection:
- Knf LabTech – Selecting a Rotary Vane Vacuum Pump: [invalid URL removed] This resource from a leading manufacturer provides guidance on choosing an appropriate vacuum pump for rotary evaporation based on specific needs.
Overall Benefits:
- Royal Society of Chemistry – Rotary Evaporation: [invalid URL removed] The Royal Society of Chemistry emphasizes how mastering rotary evaporation techniques can lead to increased efficiency and improved research outcomes.
Phind V9’s output:
A key feature of PhindV9 is its integration of source attribution on its interface: everytime it provides you with an output, it automatically shares the list of sources that it used to provide you with said output (see image below), making it easier to fact-check everything it gives you.
Discussion:
When evaluating large language models for informative tasks, phindv9 stands out for its ability to provide users with direct links to the sources it uses in every response. This allows users to quickly verify the information and delve deeper if desired. While other models, like Gemini, may offer general links to resources (most of which had invalid URLs in our example), Phind V9’s focus on direct source attribution streamlines the research process. It’s important to note that not all models can provide this functionality, as evidenced by ChatGPT’s current limitations in incorporating real-world data.
Conclusion:
Lack of Nuance: The life sciences field thrives on precision and nuance. Technical vocabulary, complex concepts, and ethical considerations demand a depth of understanding that AI models, despite their vast data stores, often lack. This can lead to inaccurate or misleading copy, a potentially dangerous pitfall in a field where clarity is paramount.
The Human Touch: The best life science copywriting resonates with the audience. It speaks to their fears, hopes, and aspirations in a way that is both informative and emotionally engaging. This is where the human touch remains irreplaceable. AI can generate text, but it cannot infuse it with the empathy and understanding that a skilled copywriter can.
Creative Roadblocks: While AI excels at generating standard content formats, it often struggles with the truly creative. Brainstorming unique ideas, crafting compelling narratives, and breaking through creative roadblocks are still the domain of human ingenuity. AI can be a valuable tool in the process, but it is not a substitute for the human imagination.
Time needed to generate a good prompt: While ALMs offer the potential to save time on writing, using them effectively often requires some back-and-forth. You might need to refine your prompts and evaluate the outputs several times. This iterative process can be valuable, but consider the time investment. Ultimately, the question is this: is it more efficient to create a detailed prompt to get the desired results from the ALM, or to write the entire piece yourself?
