SUNY AI Task Force

State your top 3 reasons for why SUNY should care about the AI wave from an education perspective?

1. AI stands to disrupt the social value of knowledge, education and educational credentials. SUNY will best serve the people of New York by being proactive and strategic in the conversation about what scholarship and college education mean and what they are worth.
2. AI stands to disrupt longstanding pedagogical practices in higher education–with opportunities and dangers. For example, the panic around using AI to cheat reveals lousy pedagogies which are past-due for replacement. But the prospect of students seriously questioning whether it’s worth learning to write is harmful to an educational environment.
3. Ai is distorting the grant funding space. SUNY can bring in more money by responding strategically. At the same time, we need to minimize damage to our scholarly capacity from this hype cycle. For example, there is pressure for people to repackage themselves as “AI experts,” and the AI wave combined with increasing pressure to bring in grants is pressuring fields to focus on questions which might not be the most important.

What should be the end vision/goal for SUNY as a leader in AI education?

1. Put social questions, not technological questions, at the center. This should include social visioning around the kind of society we want and how our research contributes to conceptualizing and enacting those futures.
2. We should actively and continuously participate in shaping the social role of universities in the age of AI.
3. We should be world leaders in pedagogy and scholarly practice, using the disruption from AI to emphasize the human (and sociotechnical) relationships and practices that constitute learning and research.

What should SUNY’s top 3 objectives be in any effort towards advancing AI education (or perhaps, in fundamentally transforming SUNY’s education system in response to the AI era)?

1. Get control of the narrative. AI hype, and the particular manifestations of AI technology has been completely shaped by self-interested tech companies. We should be de-emphasizing the inevitability and exceptionalism of AI. For example, rental scooter/bike companies deployed their services in cities all over the world while flouting local regulations. They tried to use narratives of exceptionalism and inevitability to shield themselves from regulation. This was ultimately only partially successful; cities such as Paris and Rome now have control over scooter companies.
2. Support more interdisciplinary research, particularly across computing fields and social sciences. We need research which is strong both in its technical merits/innovation, and in conceptualizing its social significance.
3. Emphasize “broader impacts:” the social value of scholarship, education, and educational institutions. This means more research-practice partnerships, more community-based research, more civic engagement in research, and more technology transfer/scholar-activism which realizes the social possibilities of research.

Nationally, or globally, which institutions and AI programs would you consider are a benchmark for SUNY to emulate? Please prioritize to top 3.

1. Stanford University
2. Arizona State University (ASU)
3. Massachusetts Institute of Technology (MIT)

Why do you consider them the benchmark?

1. (Stanford University) The interdisciplinary diffusion of computing in teaching and research. Specifically, Stanford’s attention to human-computer interaction and design. SUNY (or at least UB, where I work) is quite sioled. Interdisciplinary research is relatively uncommon, and it’s quite difficult to create research groups or offer courses which draw from across multiple departments. This is partly structural and partly due to cultural norms.
2. ASU and SUNY are already long-standing leaders in distance education, creating social mobility, and the civic role of state universities. We should make sure that as we reshape ourselves in light of AI, we remain leaders in these areas.
3. MIT is incredibly good at shaping public narratives around its research and innovations (sometimes this is harmful, for example MIT Media Lab scandals in recent years). SUNY should be more visible in the public sphere, both with flashy innovation and with solid, thoughtful scholarship which is received as definitive and authoritative.

What are the biggest gaps at your institution’s (or SUNY’s to the extent that you are aware) portfolio of AI educational offerings (or approach to AI education) when compared to the best in the nation/world? Please prioritize to top 3.

1. Interdisciplinary coursework bridging CS and social sciences. Or courses within either of these domains which have well-informed framings of the other.
2. Bringing our research to market/social impact, emphasizing design and real-world significance (broader impacts).

What are your recommendations on how to bridge these gaps. Please prioritize to top 3.

Pilot and prototype interdisciplinary courses. Create incentives for students–particularly doctoral students–to enroll in them. Support interdisciplinary doctoral research. (The incentives are more difficult for faculty, and SUNY has less local leverage on prioritizing interdisciplinary research.)

What specific opportunities do you see to advance/accelerate AI education across SUNY by pooling/leveraging resources across our institutions? Please prioritize to top 3.

Share organizational models that are working. Share useful infrastructure. For example, in a course I am teaching next semester, we will be designing educational products which utilize locally-hosted LLMs (based on Meta’s open-source llama models). This could be an infrastructural service made available SUNY-wide. We have excellent research infrastructure, but it’s not very inviting and IT people are often not excited about trying new things. I end up hosting a number of my computing education research projects offsite because of the red tape.

What operational/business model/policy/cultural changes may be required to realize these opportunities? Please prioritize to top 3.

1. Create interdisciplinary courses/programs/centers across computing fields and social sciences, with incentives particularly for doctoral student participation.
2. Invest in marketing/public understanding of SUNY research, through which we assertively shape the narrative around AI.
3. Rapidly prototype models of courses, programs, policies, and technology services, with support for dissemination of what’s working.

What do you think are the best approaches to leveraging the best of what emerging AI tools based on LLMs, such as ChatGPT, can offer towards enhancing teaching and learning outcomes without compromising integrity/trust?

Our conceptualizations of learning outcomes, as well as assessment of learning outcomes, are underdeveloped in higher education (compared, for example, to K12 education). Capacity-building in this area would be helpful, so that faculty and departments can be more thoughtful about what kinds of teaching and learning they actually value.

What are the top 3 areas of learning and teaching that you believe will most benefit from using effective AI tools (if available)?

1. Knowledge-building: Synthesis of existing knowledge in the service of specific new theoretical or practical questions. Students have so much time learning the mechanics of literature reviews, and navigating the ocean of scholarship that it’s hard to focus on the goals of what we’re doing with all this knowledge. In a sense, undergraduate education could start to resemble graduate education.
2. Responsive, multi-modal, accessible media which adapts to the learner. Which concepts or fields are unnecessarily difficult?
3. Disruption of assessment practices emphasizing competition. If we come to terms that it’s impossible to stop cheating, how will we reframe coursework to remove situations where cheating would be relevant?