An increasingly large portion of industries find the value in using ChatGPT or other AI to assist with tasks. Accuracy is not always perfect, but the same could be said about humans. The primary benefit of AI is not improved accuracy over humans. AI can instead do things humans cannot, such as near instantly sort through, interpret, and rank volumes of information; which has countless uses related to college search and selection.
In my trials, when ChaptGPT searches the web, it provides links to the sources. While ChatGPT can use Reddit, I’ve never seen a Reddit post among those sources for my college inquiries. For example, in the college list at the bottom of this post; the referenced sources were websites of the listed colleges (and websites of some colleges that are not listed), various books, US News, Niche, College Transitions, Best Colleges . com, College Vine, the National Association of College Attorneys, and over a dozen other sources… but not Reddit, CollegeConfidential, or similar forum.
Lindagaf’s examples were subjective questions such as friendliest colleges or colleges where most students believe in UFOs. If you use objective questions, you can more accurate answers. As an example of objective measures, I tried the following prompt:
Make a table showing a ranked list of colleges based on the following 10 criteria.
Criteria 1: Majors – Importance = Very High
College must offer both engineering majors and computer science related major. Ideally the college would offer both an electrical engineering major and computer science major, or an EECS a major. A general major with concentrationin EE or similar is also acceptable. Show available EE and CS majors as a column in the table.
Criteria 2: Selectivity – Importance = High
The student prioritizes having a high concentration of high achieving students. The more selective the better. However, the student also wants to apply to some “target” and “safety” colleges where he may have a better chance of being accepted, so it is okay to display colleges that are not extremely selective in the table as a lower ranking. Include columns for acceptance rate and SAT score in the table.
Criteria 3: Cost – Importance = Moderate
The student is a New York State resident. Parents income is likely too high to qualify for need based FA. Student may be eligible for merit scholarships (4.0 GPA, 1500 SAT). Cost is not critical, but may influence decisions. Show cost for NYS resident as a column in table.
Criteria 4: Location – Importantce = Moderate
The student would prefer a college where he could regularly travel home to Albany NY. Ideally the college would be within in 3 hour drive from Albany. A 5 hour drive is also good. Distance is not a dealbreaker, but closer is better. Show driving distance to Albany, NY as a column in the table.
Criteria 5: Weather – Importance = Moderate
The student would prefer a moderate climate that supports outdoor weather, where both summer and winter high are as close to 70F as possible, with over 90% of days having no rain. This isn’t a dealbreaker, but all things being equal, better weather is preferable. Show median winter high temperature and % sunny days as columns in table.
Criteria 6: Jobs / Located Near Tech Hub – Importance = Moderate
The student would prefer a college located near a tech hub, like Silicon Valley, to improve odds of tech internships and jobs. Include as column in table.
Criteria 7: College Size – Importance = Moderate
Ideally the college would have ~5000 undergrads. 2000 to 10,000 is okay. <2000 may be too small. >10000 may be too big. List undergrad size as a column in the table.
Criteria 8: Prestige – Importance = Moderate
List prestige as a column in table. Interpret “prestige” as you feel is most appropriate. List a related column in table.
Criteria 9: Football Team-- Importance = Low
The student would prfer a college that has a football team, with good shcool spirit. Ideally the football team would be a Div I team with good win/loss record and rivalry with neighboring colleges. Div II / III are acceptable, but less preferable. List related column in table
Criteria 10: Good Resources – Importance = Medium
The student would prefer a college that has good campus resources, particualrly in related to engineering, as well as good spending on students. List endowmnet per student as a column in table.
The output was the following. One can quibble with things like while RPI has both Div I and Div III teams, football is Div III, not Div I; but I think overall accuracy was good. I also think it’s a reasonable list of colleges for the listed criteria. This is an arbitrary example. One could ask ChatGPT to combine dozens of criteria rather than just the 10 as listed, or include criteria for which results may be less commonly known, leading to outputs that would be near impossible for a person to manually discover on their own.