Resea ch Pape
Recommended ci a ion: Hingle, A., & Joh i, A. (2025). Mapping S uden s’ AI
Li e acy F aming and Lea ning h ough Re lec i e Jou nals. In Kangaslampi, R.,
Langie, G., Jä inen, H.-M., & Nagy, B. (Eds.), SEFI 53 d Annual Con e ence.
Eu opean Socie y o Enginee ing Educa ion (SEFI), Tampe e, Finland. DOI:
10.5281/zenodo.17631356.
This Con e ence Pape is b ough o you o open access by he 53 d Annual Con e ence
o he Eu opean Socie y o Enginee ing Educa ion (SEFI) a Tampe e Uni e si y in
Tampe e, Finland. This wo k is licensed unde a C ea i e Commons
A ibu ion-NonComme cial-Sha e Alike 4.0 In e na ional License.
Mapping S uden s’ AI Li e acy F aming and Lea ning h ough
Re lec i e Jou nals
A Hingle a,
1
, A Joh i b
a Geo ge Mason Uni e si y, Fai ax VA, USA, h ps://o cid.o g/0000-0002-6178-1256
b Geo ge Mason Uni e si y, Fai ax VA, USA, h ps://o cid.o g/0000-0001-9018-7574
Con e ence Key A eas: Digi al ools and AI in enginee ing educa ion; Building he
capaci y and s eng hening he educa ional compe encies o enginee ing educa o s
Keywo ds: AI li e acy; AI educa ion; e lec i e jou nals; sel - e lec ion
ABSTRACT
This esea ch pape p esen s a s udy o unde g adua e echnology s uden s’ sel -
e lec i e lea ning abou a i icial in elligence (AI). Resea ch on AI li e acy p oposes
ha lea ne s mus de elop i e compe encies associa ed wi h AI: awa eness,
knowledge, applica ion, e alua ion, and de elopmen . I is impo an o unde s and
wha , how, and why s uden s lea n abou AI so o mal ins uc ion can be e suppo
hei lea ning. We conduc ed a e lec i e jou nal s udy whe e s uden s desc ibed
hei in e ac ions wi h AI each week. Da a was collec ed o e six weeks and
analyzed using an eme gen in e p e i e p ocess. We ound ha he pa icipan s
we e awa e o AI, exp essed opinions on hei u u e use o AI skills, and con eyed
con lic ed eelings abou de eloping deep AI expe ise. They also desc ibed e hical
conce ns wi h AI use and saw hemsel es as in e media ies o knowledge o iends
and amily. We p esen he implica ions o his s udy and p opose ideas o u u e
wo k in his a ea.
1
A Hingle
[email p o ec ed]
1 INTRODUCTION
Gi en he ex ensi e impac o a i icial in elligence (AI) ools, echniques, and
sys ems on e e yday li e, he e is a g owing call o ad ancing AI li e acy among
s uden s and he la ge popula ion mo e gene ally (Long & Mage ko, 2020; Tenó io
e al., 2023). Howe e , de eloping hese compe encies is challenging gi en he
complexi y o AI as a domain and because lea ne s b ing a as ange o p io
knowledge, skills, and expe iences o hei expe iences wi h AI (Ho nbe ge e al.,
2023) (Heyde & Posegga, 2021). We conduc ed a jou nal s udy o examine how
s uden s lea n abou AI and how hey en ision hei lea ning, o mally and in o mally
ou side he class oom, will a ec hei cu en and u u e engagemen wi h AI
(Ba on, 2004). The esea ch ques ions ha guided his wo k we e: 1) how do
unde g adua e echnology s uden s’ e lec ions on lea ning and in e ac ing wi h AI
align wi h es ablished componen s o AI li e acy om p io esea ch (e alua e, use,
c ea e, and e hically na iga e AI) and 2) how do s uden s en ision and alue he ole
o AI in hei lea ning and u u e? This pape builds upon he me hodological
ounda ion es ablished in p io wo k (Hingle & Joh i, 2024). We ex end he scope o
analysis in ou p io wo k o assess s uden s’ p og ess owa d di e en AI
compe encies h ough hei e lec ions. By doing so, we hope o unde s and mo e
abou how s uden s build li e acy abou AI bo h in and beyond he class oom.
2 PRIOR WORK: FRAMING AI LITERACY
Resea ch on de eloping AI li e acy is a g owing a ea o schola ship and
in e en ions in he class oom (Long & Mage ko, 2020; Ng e al., 2021; Tenó io e
al., 2023). A ecen sys ema ic e iew ou lines he signi ican e o s in AI li e acy
o e he pas i e yea s (Alma a i e al., 2024) and, based on an analysis o o he
a icles, concep ualizes AI li e acy as he abili y o ecognize, e alua e, use, c ea e
AI, and he capabili y o na iga e AI applica ions e hically. To achie e momen um in
imp o ing AI li e acy, he e iew ecommends a ious ins uc ional ac i i ies,
including in oducing e e yday examples and supplemen ing ha wi h knowledge o
AI ools and applica ions and awa eness. Finally, educa ion abou he limi a ions o
AI, including e hical conce ns such as ai ness, accoun abili y, anspa ency, e hics,
and sa e y, and unde s anding AI’s ole in ou wo ld and i s impac on socie y, bo h
p esen and u u e, is conside ed an essen ial elemen , oo.
Recognizing o becoming awa e o AI is a undamen al i s s ep in building AI
li e acy. Howe e , implemen ing o mal measu es o aise AI awa eness is
challenging because people o en in e ac wi h AI in unno iced ways in hei daily
li es. Iden i ying AI’s sub le ole ac oss socie y is c ucial o de eloping an in o med
and c i ically engaged popula ion ha can in e ac wi h AI echnologies esponsibly.
To his end, Koenig (2020) p esen s a s udy ha engages w i ing s uden s in
e lec ing on hei daily in e ac ions wi h algo i hms h ough media jou nals. The
analysis sugges s ha s uden s became mo e awa e o he unde lying mechanisms
by documen ing and analyzing hei engagemen s wi h pla o ms like Facebook,
Amazon, and Google. W i ing he jou nals encou aged s uden s o examine pa e ns
and ques ion hei ole in main aining hem. The s udy sugges s ha while s uden s
ini ially unde s and algo i hmic p ocesses a a basic le el, e lec i e jou naling
p omp s hem o de elop a mo e nuanced, c i ical awa eness. The jou nals also
highligh ed ha no e e y lea ne engaging wi h AI wan s o de elop AI. Some y o
be e unde s and he sys ems hey use daily and be in o med use s.
To be consis en wi h p io wo k, his s udy si ua es AI li e acy in e ms o lea ne s’
abili y o ecognize, e alua e, and use AI and hei ocus on e hical o social aspec s.
Howe e , i explo es he de elopmen o hese cons uc s h ough lea ne s’ sel -
e lec ions. F om a me hodological pe spec i e, his s udy uses jou nals o allow
s uden s o desc ibe hei e e yday in e ac ions wi h a ocus on he componen s o AI
li e acy. The jou nals also enable pa icipan s o build on hei expe iences each
week and build on connec ions be ween in e ac ions hey discussed ac oss he
weeks. The easily explainable o appa en engagemen s o en eme ge in he i s o
second jou nal, and s uden s eel he need o do mo e in subsequen jou nals.
3 METHODS
3.1 Da a Collec ion h ough Re lec i e Jou nals
Re lec i e ac i i ies a e ecognized ac oss cu icula, allowing lea ne s o g apple wi h
knowledge and li ed expe iences o c ea e meaning, in eg a e hough s, and build a
deepe pe spec i e (B ockbank & McGill, 1998; Bolge e al., 2003). They allow
lea ne s o conside decisions, ques ions, and al e na i es and syn hesize he
knowledge (Koenig, 2020). Jou nal-based da a collec ion is commonly used in
educa ional esea ch o cap u e s uden s’ ongoing lea ning p ac ices (A nd & Rose,
2023). Ou app oach is buil on p e ious s udies using e lec ion o unde s and sel -
egula ed lea ning (Schmi z & Schmid , 2011; Ro h e al., 2016).
The s udy pa icipan s submi ed e lec ions h ough a e lec i e jou nal en y each
week o six consecu i e weeks (ea ly Oc obe o mid-No embe 2023) on hei
in e ac ions wi h AI. Pa icipan s we e asked o answe an open-ended p omp ha
encou aged hem o e lec on any aspec o hei engagemen s – wha hey did, how
hey in e ac ed, whe e i occu ed, and wha imp ession he in e ac ion le on hem.
The p omp gi en o s uden s was:
In he p e ious week, documen whe e and wha you ha e no iced men ioned
abou AI: a you campus o ou side o you campus? Did he desc ip ion o AI
lea e you wi h a posi i e, nega i e, o neu al imp ession? Why o why no ?
The p omp was designed o be b oad o cap u e a ious AI- ela ed expe iences and
e lec ions while a oiding guiding s uden s’ hinking o AI in a speci ic. B oad
ca ego ies o AI li e acy we e iden i ied du ing he de elopmen o he e lec i e
jou nal and he p omp s om p io e iews (Long & Mage ko, 2020; Tenó io e al.,
2023; Ng e al., 2021). These included bo h echnical and non- echnical
compe encies, in addi ion o implemen a ions ac oss di e en use cases.
3.2 Pa icipan s
In Sep embe 2023, wen y- wo (22) s uden s we e ec ui ed om a cou se designed
o discuss he impac o echnology wo ldwide o pa icipa e in his s udy. The
pa icipan ’s sel -desc ibed gende a io was 11 emale: 11 male. The mean age o
he pa icipan s was 23 yea s ( ange: 19-32 yea s). Mos pa icipan s we e ull- ime,
pu suing a BS in In o ma ion Technology wi h a ious specializa ions, including
de elopmen , cybe secu i y, da abases, and heal h in o ma ion echnology. The
pa icipan s comple ed an a e age o 80 cou se c edi s ( ange: 40-121). Pa icipan s
we e gi en a $48 gi ca d o comple e he six jou nal en ies. No cou se c edi was
o e ed o pa icipa ing.
3.3 Da a Analysis P ocedu es
We used a hyb id app oach o analyze he 22 collec ed e lec i e jou nal en ies,
which comp ised six en ies each o 132 e lec i e jou nal en ies. Fi s , we used
eme gen coding o cap u e all pa icipan ideas due o he a ied con en o he
jou nals, and second, we used he AI li e acy amewo k (awa e-knowledge-use-
e alua e-c ea e) o p o ide s uc u e o he analysis. The app oach was simila o
Koenig’s s udy on using jou nals o algo i hmic li e acy awa eness (Koenig, 2020).
A e collec ing he da a, wo esea che s i s ead h ough he jou nals and
highligh ed in e es ing jou nal en ies om which a se o ini ial codes was gene a ed.
Re iewing hese ini ial codes, he esea che s ecognized simila ideas o hose
highligh ed in ea lie li e a u e e iews - a ocus on awa eness, knowledge, usage,
e alua ion, and he e hical implica ions o AI use, in addi ion o de eloping AI. Using
he i e cons uc s, he esea che s e-coded he da a wi h he s uc u e as a guide,
ensu ing ha any in e es ing codes ha did no i in we e e ained. These codes
we e hen g ouped in o hemes, which we e e iewed and inalized.
4 FINDINGS
Themes ac oss he jou nal en ies sugges ha s uden s we e awa e and la gely
cau ious o AI’s p esence in hei daily li es. They equen ly connec ed hei cou se
lea ning wi h expe iences ou side he class oom. Many pa icipan s wen beyond
simple awa eness and desc ibed hei hopes and ea s o AI and hei expec a ions
o wha li e acy could be, aligning wi h he mo e ad anced elemen s o AI li e acy. In
his sec ion, we p esen he indings om he e lec i e jou nal en ies, using he
ca ego ies o awa eness, knowledge, applica ion, na iga ing e hically, and
de elopmen . Th oughou , s uden s desc ibe hei pe cep ions o he alue o
lea ning abou and using AI and he po en ial impac hey expec o hei u u e.
4.1 Awa eness o AI
Ac oss he e lec i e jou nal en ies, pa icipan s discussed hei awa eness o AI and
a le el o ecogni ion o awa eness hey belie ed gene al people should possess. In
a leas one o he jou nal en ies hey w o e, all he pa icipan s desc ibed an
expec a ion o a minimum le el o socie al unde s anding. Mos pa icipan s
desc ibed he lack o knowledge as posing as a pe sonal and p o essional ba ie .
As a s uden , I hink lea ning abou AI is no a choice. We ha e o be com o able
wi h AI i we wan o do well in his ield. Bu I am no su e i i is he same o o he
majo s. I hink hey need o know wha i is, bu I don’ hink hey eally need o lea n
how o p og am o do machine lea ning. [P1 Week 3]
Pa icipan 1 highligh ed ha hough he basic le el o unde s anding will di e
depending on he ield o s udy, awa eness is s ill impo an . Se e al pa icipan s,
such as Pa icipan 9, emphasized he unde s anding componen o ex end beyond
lea ne s in he class oom o a gene al awa eness o AI o people:
I was a li le conce ned a e [class discussion] because I ha e amily ha I hink
may be in dange because AI can do hei wo k so much as e and cheape . I is
ha d o alk o hem abou hese issues because hey don’ know much abou AI o
echnology a all. [P9 Week 5]
Pa icipan 9 highligh ed he challenge o engaging in discussions wi h indi iduals
who lack a undamen al awa eness o AI concep s, le alone unde s anding he
de ails, and ha hey may be disad an aged in how hey can adap .
4.2 Knowledge o AI
While awa eness e e s o a gene al ecogni ion o AI’s exis ence and amilia i y wi h
whe e AI is being used, knowledge e e s o AI concep s, echniques, and skills.
Mos pa icipan s dis inguished be ween a high-le el o gene al unde s anding o AI
concep s and a mo e echnical unde s anding o building and de eloping AI models.
They no ed ha a mac o-le el unde s anding encompasses knowing wha AI is, how
i can be applied ac oss a ious domains, and i s po en ial socie al impac s. This
ype o unde s anding is c ucial o making in o med decisions abou AI’s use and
ecognizing i s b oade implica ions. Pa icipan 7 desc ibed wan ing o lea n some o
he echnical componen s bu mos ly being in e es ed in how he sys ems wo k:
I hough [a icle abou AI job changes] was in e es ing because i highligh s how
impo an AI will be going o wa d o e e yone o ha e some unde s anding o . I
pe sonally don’ in end o be a de elope , bu I s ill ind he echnical side o hese
sys ems o be e y in e es ing. I hink ha ing deepe knowledge, e en i I don’ wan
o build AI, should gi e me an ad an age o e o he s. [P7 Week 3]
Suppo ing his dis inc ion, some s uden s highligh ed ha hough wha hey
encoun e ed was highly echnical, hey pe se e ed as hey ound his one a enue o
isola e he concep s hey would likely need in he u u e. Pa icipan 17 desc ibed
coming ac oss wha seemed o hem like an ad anced opic bu app oached i
anyway wi h an open mind o lea ning:
I no iced ano he semina om he s a s depa men i led [semina i le] abou using
AI models o unde s and complex p oblems a ound ene gy and he en i onmen .
The name was in imida ing, bu I s ill a ended. I s ill ha e a lo o lea n bu I wan o
do his kind o wo k in he u u e so I am happy o lis en. [P17 Week 5]
O e all, pa icipan s dis inguished be ween a highe -le el mac o- hinking and
lea ning abou AI compa ed o lea ning o c ea e.
4.3 Using AI
As expec ed, many pa icipan s desc ibed he di e en AI uses hey engaged wi h.
These included so wa e, sys ems, and ools as a se ice o accomplish a ask. In
addi ion o desc ibing how hey used AI, many pa icipan s a icula ed ha using AI is
highly ela ed o hei lea ning goals. They did so in ending o enhance hei
compe encies wi h ools hey assumed would be use ul in he u u e. They also o en
exp essed a b oad desi e o s ay knowledgeable on cu en ools and sys ems in a
apidly e ol ing ield. Some pa icipan s, such as Pa icipan 20, desc ibed hei
goals in lea ning speci ic skills ha would be use ul o hem la e in hei ca ee s:
In my [cybe secu i y class], we alked abou Splunk and he di e en AI capabili ies
ha a e a ailable in he so wa e and how hey make hings easie o
adminis a o s (like he anomaly de ec ion ea u es). Those kinds o hands-on
classes eel especially use ul and a e why I wan ed o ake elec i es ha alk abou
AI in hem. [P20 Week 2]
Simila o he discussions on a gene al unde s anding o AI, some pa icipan s
desc ibed hei ope a ional in en ions o unde s and and e icien ly use he sys ems.
Pa icipan 22 desc ibed how ec ui men applican acking sys ems (ATS) se e as
an example o applica ions e e yone would likely need o be use ul:
I don’ wan o go in o p og amming, bu I hink I will p obably use some kind o AI in
my wo k ei he way, so knowing how hey wo k is impo an . One o my g oup
membe s b ough up how mos companies use ATS so wa e o sc een esumes,
and e en he e, knowing how i wo ks helps you ge h ough he p ocess. [P22
Week 6]
Gene a i e AI (GenAI) eme ged in hese discussions o en, and he e was a ie y in
he b ead h o an icipa ed use.
I al eady use Cha GPT e e y day, and I ha e ull in en ion o make he mos ou o
he addi ional powe i p o ides o a mo e han casual echnology use like me.
Ge ing be e a p omp ing is some hing I plan o do soone han la e . [P8 Week 6]
4.4 E alua e AI
Though o en desc ibed as wo sepa a e cons uc s, pa icipan s equen ly
discussed e hical and socie al implica ions in conce wi h c i ically e alua ing AI. The
pa icipan s p ima ily indica ed un amilia i y wi h echnology e hics, wi h many
encoun e ing he opic o he i s ime in he cou se. None heless, as Pa icipan 22
highligh s, hey exp essed a s ong in e es in u he explo ing he subjec om o he
in o med pe spec i es:
I ha e s a ed ollowing people and g oups on LinkedIn ha alk abou he e hics o
using da a in di e en ways. [P22 Week 6]
Pa icipan s pe sis en ly aised he issue o da a bias, likely e lec ing i s co e age in
hei cou sewo k. No ably, hey demons a ed a deepe le el o c i ical hinking on
his issue. As Pa icipan 9 desc ibed, da a bias can a ec he en i e p ocess om
collec ion o any decisions made as a esul :
The e a e so many issues wi h he bias and ypes o da a ha a e al eady in he
sys em, I don’ know i hese can e e eally be ixed. The bias in he da a is buil in
om he s a and i is used h oughou he p ocess. So, is he e an easy way o
e en go abou wo king wi h his ype o da a? I eally don’ hink so. [P9 Week 2]
Pa icipan s also connec ed he implica ions o bias in da a collec ion and i s
u iliza ion in models, highligh ing how hese ac o s can p ese e disc imina o y
p ocesses and p ocedu es.
I no p ope ly ained, AI can pe pe ua e and ampli y exis ing biases in da a, leading
o disc imina o y ou comes, especially agains ma ginalized g oups. [P2 Week 2]
Al hough he pa icipan s we e p imed o hese discussions h ough cou se
eadings, hei choice o add ess hese opics in a jou nal, whe e hey had he
eedom o discuss any hing, is a posi i e inding o ans e ing lea ning beyond he
cou se ma e ial. Pa icipan s g a i a ed owa ds opics on AI's social impac , such as
bias o e hics. In he analysis, pa icipan s o en began desc ibing o he conce ns
associa ed wi h da a-d i en decision-making, such as hose o powe dynamics,
su eillance, and sa e y, bu could no speci ically desc ibe hem.
4.5 De eloping AI
While including he de elopmen o AI o he use o machine lea ning echniques
speci ically in AI li e acy measu es is con es ed (Ca olus e al., 2023), pa icipan s
o en desc ibed hei in en ions o lea ning abou AI h ough his aming. This may
be because all he pa icipan s we e in a echnical p og am whe e hey we e
equi ed o ake p og amming, algo i hms, and machine lea ning cou ses. Mos
pa icipan s desc ibed he easibili y o applying hese skills in he u u e. They
desc ibed he knowledge and beha io s hey acqui ed as highly ele an and
ans e able o academic, p o essional, and en ep eneu ial con ex s. Some
pa icipan s also hough o using GenAI simila ly:
I was eally in e es ing o ead abou how [Copilo ] is being included in he Gi Hub
en i onmen and wha his could po en ially mean abou gene a ing code o helping
people w i e code. I hink his is a wo hwhile idea o wo k on because coding can
be di icul , and we wan as many people o unde s and how o code. I ind w i ing
code o be in imida ing, which is why I don’ plan on becoming a p og amme , so
ha ing help like his makes me eel like I can use i when I wan . [P11 Week 5]
Finally, some pa icipan s exp essed a desi e o enhance hei AI li e acy o a le el
su icien o pu suing en ep eneu ial en u es and building a business:
Bu wi h AI being so good a looking o pa e ns, maybe i is be e ha [ag icul u al
sys ems, such as hose used o e icien ly wa e c ops] a e aken ca e o by AI. I
am eally in e es ed in how hese sys ems wo k, and a e wo king o a ew yea s, I
wan o be able o c ea e my own business ha uses AI. [P20 Week 5]
5 DISCUSSION AND IMPLICATIONS
The indings p esen se e al conside a ions ega ding bo h esea ch ques ions. Fi s ,
h ough he aming o AI awa eness o knowledge, he idea o a baseline AI
unde s anding was uni e sally desc ibed by all he pa icipan s. This aligns wi h he
discussions a ound AI li e acy ha ha e eme ged o e he las decade (D uga e al.,
2019; Long & Mage ko, 2020; Tou e zky e al., 2019). Some pa icipan s a gued ha
his would ensu e e e yone could make he mos o echnological ad ances, bu
o he s ook a p o ec i e s ance ega ding li elihoods and consume p o ec ion. The
pa icipan s p esen ed a case o a echnologically awa e popula ion a a gene al
le el as a minimum. Howe e , de ining wha he minimum should be is challenging.
One majo challenge is ha lea ning abou AI basics can be di icul i he lea ne
does no unde s and he p e equisi es. Especially when pa icipan s desc ibed
gene al AI li e acy, when aking he pe spec i e o a iend o amily membe , hey
b ough o a en ion ha hey we e comple ely unawa e o AI. The disconnec
be ween how common AI ools appea in e e yday spaces was los o hem. I
equi ed he pa icipan s o s ep in as in e media ies, en o cing hei unde s anding
and conce ns abou AI wi h o he s.
Access o esou ces, bo h he abili y o pay o an AI subsc ip ion, educa ion on he
opic, o men o ship may exace ba e he issue. Celik p esen s he digi al di ide
(unequal access o echnologies) and compu a ional hinking as de e minan s o AI
li e acy (Celik, 2023), and his is e lec ed in some esponses in his s udy. Especially
when pa icipan s desc ibed a gene alized le el o li e acy, when aking he
pe spec i e o a iend o a amily membe , hey b ough o a en ion ha hey we e
comple ely unawa e o AI. The disconnec be ween how common AI ools show up in
e e yday spaces was los o hem and equi ed he pa icipan s o s ep in as
in e media ies, a ole ha p o ides hem wi h an oppo uni y o lea n by eaching
(Palinsca & B own, 1984). S ill, i is a ole ha hey may be unp epa ed o .
Rega ding using AI, pa icipan s highligh ed how hei in e es s played a ole in
na u ally building li e acy. Pa icipan s p edominan ly conside ed lea ning abou AI
as a hobby, which could be pa o he cul u e o being in a echnology- ocused
p og am. In e es in compu ing opics can play a ole in de ining he lea ne ’s goals
(Bollin e al., 2020). Howe e , his may be di e en i hese pa icipan s we e om
non- echnical p og ams. As such, ex ending ca ee men o ing, coaching, and
ealis ic expec a ion se ing should be a pa o building AI li e acy among s uden s.
In his s udy, mos pa icipan s implied ha hey did no ha e men o s in his ield.
Mos pa icipan s exp essed in e es in using he echnical skills hey we e
de eloping a ound AI in hei u u e ca ee s. Howe e , hey we e spli on how o do
so. This mi o ed he discussion on baseline unde s anding and he lea ne s’ goals.
Pa icipan s a icula ed di e en pe sonal decision-making laye s and hough abou
how and why hey a e lea ning abou AI. Access o in o ma ion, ce i ica ions, and
aining a e abundan , and ocusing on imp o ing one’s compe encies wi h AI
in ol es an oppo uni y cos whe e he s uden could be lea ning o doing some hing
else. Men o ship may also ha e a ec ed his in en ionali y.
F om he e lec ions, i was appa en ha he pa icipan s we e na iga ing he hype
o AI while being cau ious o he socie al implica ions. Some pa icipan s implied
echnosolu ionis hinking o using AI o sol e e e y p oblem. Howe e , mos
pa icipan s exp essed unce ain y abou how he skills hey we e de eloping would
be used, e en hough hey s ill hough hey would be use ul. O he s desc ibed
pu posely aking on challenging opics in semina s and cou ses despi e no knowing
how use ul hey would be because hey hough his was hei expec a ion as
s uden s. Fo many pa icipan s, i appea ed ha he deba e ega ding he
impo ance o AI had al eady been esol ed – AI was seen as ubiqui ous and c i ical
– and hey we e me ely esponding o his es ablished consensus.
6 LIMITATIONS AND FUTURE WORK
The s uden popula ion olun ee ed o pa icipa e and hus sel -selec ed, and all
majo ed in in o ma ion echnology, which limi s he gene aliza ion o his s udy. The
eme gen analysis p esen s some indings in de ining hemes, bu his s udy did no
e alua e he co ela ion be ween he quan i a i e i ems. We encou age u he
explo a ion wi h explainabili y as a cen al ocus. In his wo k, we do no explo e any
da a om ollow-up discussions. Howe e , u u e wo k will explo e he impac o a
pos -in e en ion discussion, ei he a an indi idual o g oup le el. This may be
excep ionally use ul o building AI li e acy and awa eness. Finally, he e a e
limi a ions o his wo k as AI is a as changing ield and new de elopmen s a e likely
o change s uden s’ iewpoin s and unde s anding equen ly.
7 ACKNOWLEDGEMENTS
This wo k is pa ly suppo ed by US NSF Awa ds 2319137, 1954556, and a
USDA/NIFA Awa d 2021-67021-35329. A esea ch g an om Geo ge Mason
Uni e si y also suppo s his wo k. Any opinions, indings, conclusions, o
ecommenda ions exp essed in his ma e ial a e hose o he au ho s and do no
necessa ily e lec he iews o he unding agencies.
