Image
Deep Dive
Navigating AI Responsibly: A Practical Guide from the Pace Library
ItÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s our new best friend!
ItÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s the end of critical thought!
It will destroy/revolutionize education!
Many if not most of us are grappling with understanding and learning a suddenly pervasive technology: generative AI (GenAI). Like most new technologies, GenAI carries a load of anxieties along with its benefits, presenting not only skills issues but also wider ethical questions.
Though the death of writing appears to have been exaggerated, there is still plenty to be concerned about during this rapid adoption. Is it possible to use GenAI in a way that feels safe and principled?
Here are a few things you might worry about when using ChatGPT, Claude, Gemini, or any of the array of AI models, and how you might adjust your practices.
Privacy
Since the advent of GenAI tools, privacy has been an issue. Any info that you put into a GenAI toolÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”including the content you create, like prompts or material you upload to work onÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”can theoretically be used to train the AI. If it trains on your work, your work may come out in another userÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s output.
Big AI companies usually claim that user information is not used as training material, but their privacy policies and terms of use say otherwise.
An even more obvious privacy problem comes from the fact that AI companies can collect your identity and contact data, IP address (which indicates your location), device and network information, and possibly other information as available to them.
What can I do to safeguard my privacy?
Better safe than sorry. You can proactively opt out of having your data kept and possibly used by an AI company by finding the opt-out process. (Not all US states have opt-out requirements for companies that gather personal data, but enough of them do that there should be a mechanism.) The companies donÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t make it easy to locate, but it is usually in the privacy policy or the terms of use.
Environmental Cost
Generative AI is extremely sophisticated and powerful, and it requires sophisticated and powerful computers to run it. These, in turn, demand enormous amounts of energy and water.
It has been estimated by that every prompt entered into a GenAI tool consumes about a bottleÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s worth of water. ThatÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s not a lot, but itÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s 10 times as much as a Google search, and over the course of a big project, you could end up using a truckload of bottles.
What can I do to reduce AI-induced waste?
Abandoning AI isnÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t the answer. Even the greenest among us are using resources all the timeÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”simply by being aliveÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”and itÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s possible that AI will be able to reduce our energy use in the long run. For now, while the short-term costs are high, the best thing you can do is be efficient about how you use it.
Learn how to write good prompts (you can use LinkedIn Learning through Pace ITS or review PaceÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s resources on prompting), think them out beforehand, and youÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™ll need to use fewer of them.
Loss of Skills
This is the one that probably worries us, as university affiliates, the most. WeÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™re in the business of teaching and learning; what happens when we outsource planning, writing, even drawing to AI? It seems like uniquely human abilitiesÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”critical thinking, logical planning, creativityÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”canÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t help but atrophy.
What can I do to make sure my skills stay sharp?
DonÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t panic. ChatGPT may be able to produce 500 readable words that address a topic, but truly useful content requires a lot of human intervention. AI doesnÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t do your weightlifting for you; itÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s the gym equipment that makes it easy and convenient for you to do the weightlifting yourself.
As a result, those high-level intellectual skills are still very much required to get good results out of GenAI. A prompt that produces what you want must be planned and broken down, step by step, and written carefully with attention to detail and subject-specific knowledge.
Conclusion
These arenÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t the only issues with AI, and these suggestions arenÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™t the only ways to improve your relationship with AI. But, if youÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™re a member of the Pace Community, the Pace Library can help you with specific questions, instruction, class policies, and more. Ultimately, itÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s up to each of us to balance these ethical challenges with AIÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s potential, ensuring AI is used effectively, thoughtfully, and responsibly.
For more information, check out the Pace LibraryÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s and guides to AI, or set up an appointment with a librarian in or .
Deep Dive
As artificial intelligence seeps into every facet of life, Pace scholars are working to harness the technologyÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s potential to transform teaching and research. While the road ahead is fraught with uncertainty, these Pace experts see a fairer and safer AI-driven future.
Deep Dive
ÄÐÅ®²Ù±ÆÊÓƵÈí¼þ Professor of Art Will Pappenheimer, who has long incorporated digital media and new technologies into his artwork, discusses his latest AI-influenced exhibition and the technologyÄÐÅ®²Ù±ÆÊÓƵÈí¼þ™s effects on the art world.
Deep Dive
With artificial intelligence remodeling how healthcare is researched, and delivered, Pace experts are shaping the technologyÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”and erecting the guardrailsÄÐÅ®²Ù±ÆÊÓƵÈí¼þ”driving the revolution.