The Internet Detective meets Gen AI
Kevin Bell Ed.D
Head of Higher Education and Research - ANZ at Amazon Web Services (AWS)
Gen AI is not a threat to higher education. Gen AI is a tool that can assist people in answering questions and generating text, but it is not intended to replace or compete with higher education. Instead, Gen AI can be seen as a supplement to higher education, offering an additional resource for students, educators, and professionals to use for information and knowledge.
Higher education provides students with a comprehensive education and the opportunity to develop critical thinking and problem-solving skills, as well as to gain a deep understanding of their field of study. Gen AI, on the other hand, is designed to provide quick answers to specific questions and generate text, but it does not provide the same depth of understanding and development of critical thinking skills as a higher education program.
In conclusion, Gen AI is not a threat to higher education but rather a tool that can complement and support it.
So that’s ChatGPT’s answer to the question as to whether Gen AI is a threat to Higher Education.
This is now me which I will confirm by making a tie-poh - there may be others – going forwards let's just call them proof of 'human' authorship ??
Though I am loath to say so, I have elements of agreement with the bot response above. As I have reviewed and assessed the deluge of articles and op-ed pieces over the last week and a half, I am left with two persistent thoughts.
1)?????Isn’t it great that people are thinking and talking about the purpose and, particularly assessment practices in Higher Ed.
2)?????Can Gen AI write my Go to Market strategy for 2023?
Leaving the latter aside for now, I have had the privilege to work for around 20 years in Higher Education in a variety of roles around the globe. By happenstance we had a baby (I know that’s not happenstance, more like basic biology)… but as I looked at Master’s degrees I came across one of the first hybrid / blended degrees offered Stateside – in frosty New England no less. Offered out of Marlboro College, specifically the semi-autonomous (Marlboro College) Graduate School, imagined into existence by a youthful, entrepreneurial college president named Paul LeBlanc. Since those early days, on a dial-up in rural Vermont, I was fascinated by, and worked at, producing tech enhanced student experiences at institutions offering online or hybrid programmes.
As I moved onward in my career I noted that I, and my online teams, tended to be held to higher standards and asked to meet more requirements than traditional face to face was. The more innovative the ‘product’, the more explicatory collateral we were asked to provide. No philosophical problem with that by the way: just as autonomous cars have to be a hundred times safer than ‘traditional’ drivers before they are accepted by society, so online needed to verify and validate learning outcomes while somewhere an over-tired, world weary academic trying to get 100 papers graded before dinner was assumed infallible.
Can we agree to not assume that Higher Ed is impervious to improvement, and also that tech-enabled learning is not automatically a negative. Periodic querying of long held assumptions, review of means of delivering viable outcomes at scale, particularly what we do with assessment - I support challenges to pretty much any status quo there is. Obviously excepting the one (Status Quo) who sing Rocking all Over the World.
Post Marlboro College, I followed the soon-to-be-legendary-LeBlanc to Southern New Hampshire University. When Paul took up office, SNHU (pronounced SNOO, never Schh-Noo btw) was a mid-sized evolution of what had previously been New Hampshire College (to 2001), New Hampshire College of Accounting and Commerce prior to that, and originally, the even less soundbyte-y New Hampshire School of Accounting and Secretarial Science – 25 students in 1952.
As online grew - many hoops to jump though and bars to clear (as in high jump rather than Bukowski-esque escapades). This became even more pronounced when we developed and launched the humbly titled “College for America” – a competency-based, seat-time agnostic online variant of online. For anyone not familiar, this is a model where predominantly self-directed learning drives students to competency-demonstrating assessments that they can take whenever they are ready for them. These specific elements – self-directed and seat-time-agnosticism – allowed SNHU to offer the programmes at a reduced rate even against their relatively affordable (by US standards) traditional and ‘traditional online’ offerings. During development we consulted with ETS - ‘the world's largest private non-profit educational testing and assessment organization’ who are responsible in the US for GREs, TOEIC, TOEIC and a bunch of other celebrated test acronyms. I.e. Respected. As Paul said in a working session with them: “we need the assessment for this program to be rock solid.” Given that we were effectively challenging some of the long-established tenets of Higher Education – the Credit Hour, term times, traditional instruction - we knew that we needed to precisely map assessments to what we were claiming the students would have learned. Given that we had added a lot of focus on 'softer' skills such as critical thinking and ‘creativity’ this was indeed a challenge. One we worked through.
So – and apologies for the tour round the houses – this makes me an obligate, Bloom-driven pedant when it comes to correlating assessments to outcomes and brings us back to Gen AI and what it may or may not be short-circuiting.
------------------------------------------------------------------------------------------
If an assessment is designed to assess the writing process and composite skills in drafting, editing, revising, tweaking and actually penning a narrative, then perhaps an in-class, brainstorming, guided, even handwriting-centric session would be a great activity. I personally, didn’t take a literature or English language-focused course after Whitley Bay High School (aged 17 ?), and yet in my subsequent academic career, I have written, wrote, scribed and penned, literally hundreds of essays and papers. These were intended to demonstrate my understanding of Life Science, Genetics, Risk Analysis, Education Theory, Law, Pedagogy/Andragogy, Academic Governance and Learning Analytics. None - as in Not-One of the “Learning Outcome(s)” of these ‘papers’ to my recollection was: “Demonstrate that you can write a paper.”
A current colleague of mine, focused on academic Research speculated, “Imagine someone just made a scientific breakthrough that could save lives, solve climate crisis or mitigate pandemics, yet she was terrible at writing papers and couldn’t get her work accepted by a journal editor?” In the region of 3 out of every 100 research papers submitted to prominent journals such as The Cell, Nature, and Science make it past the editor and peer review process. Most top journals have around 80% rejection rates, a common reason being “Non-compliance with the writing template.”
Wouldn't it be a shame if students actually knew and could articulate and demonstrate skills and competencies but struggle and fail because of fear of a blank sheet. Ditto researchers.
One would assume that once they make it past Education, they’ll be fine; they’ll be able to use one of those everyday AI-ML powered tools to help them with their first draft - even a sh1tty First Draft as my program lead at Penn used to call it. As in - “Give it another go, there’s so much crap here, there has to be a pony somewhere” ? Robert (Bob) Zemsky 2014-ish. Perhaps a little bit of guided practice, using these tools, in an educative environment would be helpful?
If we ignore broadly available tech tools or focus on closing Pandora's box(es) - can we say we are equipping students for subsequent lives and careers in the real world? In terms of assessment, are we testing for the actual skills we are claiming they will need? or are we simply batching assessment types for the ease and familiarity of standard (ossified?) practices? These are fundamental questions that this (Gen AI) issue is - in a good way - bringing to the fore.
----------------------------------------------------------------------------------------
We’ve been here many times, at least twice even in my short lifetime. Firstly, we the theory was to ban Calculators in class and in exam rooms as students would lose the ability to calculate fast in their head. Then - remember Y2K ! - we had ban “the Internet” from classes and tests - basically because some of it’s might not be true (as in In Encyclopedia Brittanica)...
领英推荐
Clearly, swift mental calculation is a beneficial skill to have. Did the shop-person give me the right change? Is the small apartment-sized washing powder cheaper than the economy size (my local now has that information on the shelf btw) …? With post COVID cashless-pay and supermarkets with the per-100g price on the shelf, I’m now not sure that I’m using math(s) often and the world is / has not yet ended.
As for Some of the Internet is not true – I remain indebted to some of my Undergrad – Manchester University (UK) - instructors who were, by many defintions, a bit weird. One prof. would wear different socks to see if anyone would comment – lining up British reticence to challenge authority against questions of basic sartorial decency. I think (with hindsight) they were pushing us to think critically. I recall another class topic in our Risk class - How Seatbelts Kill - Answer - they make people feel overly secure so they drive like maniacs.
During my Masters at Marlboro College (USA), The Internet Detective was an interactive, online tutorial assessing (remember it's 1999-2000), the issues of information on the Internet and the skills required to critically evaluate the veracity of a resource. For a nostalgic trek down content and low-fi pre-millennial design, check out https://library.wur.nl/infoboard/bergruimte/detective_uk/off-line/0.html I’ll wait...
In summary, looking back on prior instances of technology-encroaching-onto-academia’s turf; the world did not end. My laptop, phone and watch now all have calculators and yet still, I still stave off mental decline by estimating numbers, gas mileage (miles per gallon vs litres per 100km), random percentages and the temperature in Fahrenheit (hi American former colleagues) converting from Celsius-based living circumstances. I think most of us have mitigated the issues around use of the Internet in school-work, though a revised Internet Detective to provide guidance on TikTok, Deep Fakes and deciphering Elon Musk would be helpful if anyone wants to give that a go.
Gen AI (this is still me right-ing btw) is a fascinating tool. Many classroom (former) colleagues and peers of mine are generating fascinating activities involving students amending prompts and questions then critically analysing classmates’ outputs in Socratic Q+A. I see many gainful and gameful applications of a range of AI tools.
Plagiarism – from notes on scraps of paper to contract cheating has and will always happen. Probably the only thing that would really stop it would be to formatively assess and frequently confirm understanding in the run up to high-stakes exams which, (given formative assessment and frequent confirmation of understanding), we might not actually need…
Chat GPT is the first, certainly not the last, high profile LLM (Large Language Model) writing tool to emerge and clearly merits attention. Students need to be taught about integrity AND how to exist in a modern society with calculators (check), the Internet (check), and the next big thing (pending checks). Foreign language proficiency will be the next philosophical game changer – Will we suffer from not rote learning the past participle of “to swim” in Hungarian? Or will we be more focused on the fun and learning we'll get speaking through the translator-icom ? and diving deep into different cultures, societies and global perspectives.
It is indeed, a Brave New World out there. I (personally and professionally) want nothing more to do with Lock Downs neither COVID nor tech related.
Let’s teach and learn, analyse and discuss together.
Let’s instructionally-design our way around some of the concerns and
Let’s see what capacity this all frees up to polish and hone exceptional contemporary skills in graduating students.
My advice - fasten your seatbelts (despite the elevated Risk), and stay open to 'different' potentially being Good-Different. These tools should be looked at as potentially removing the mundane (aspects) so we can focus on the really HUMAN Value-Add parts. Experience and enjoy the AI-augmented polyglot-globetrotting journey.
Let me know what you think... doesn't have to be a six-page paper ;-)
?
Field Marketing Manager at Amazon Web Services
1 年Kevin it’s my first time coming across your writing and it’s a fun read. Hope to see more in the future! brave new world is one of my favourite books and we always see an allusion to it when the next new technology comes along. For marketers I do see a lot of utility coming out of ChatGPT, though work needs to be done to enhance accuracy. Interesting to see how this technology will progress in the coming years. And no, ChatGPT didn’t write this comment ??