When it came to finding scholarships online, typical search engines would often promote the big brand scholarships with thousands of applicants that required subscriptions to their newsletters.
However, with generative AI, the results can be personalized by refeeding the chatbot prompts to help it learn as the conversation progresses. In the instance of ChatGPT, we tried a few different prompts to see the variety and versatility of scholarships they have in their database.
We used the same prompt across every engine we tested, starting with ChatGPT. When using the general prompt, “find me scholarships as a freshman in college” the chatbot gave a general answer saying it was economically wise to find scholarships. Then it listed steps to find scholarships and different sites, like Scholly, to further find actual scholarship names. Although the tips were adequate there weren’t any titles or actual application dates that could be helpful. After a few more conversation prompts we found that the most accurate information was generated through this prompt, “I am a freshman in college, a female engineering student, and a resident of Texas. Find me specific scholarships”. This prompt allowed ChatGPT to give a list of the biggest, most popular scholarships for my specific needs. Additionally, it was helpful that they gave the eligibility, amount, and deadline for the scholarship which would generally have required more research on the website itself.
Then we tested the same prompt with perplexity, which provided a similar answer in a different format. They described the eligibility and deadline but did not describe the amount for every scholarship offered. However, perplexity informed us of when approximately the application opened. Which is information that is typically more difficult to find than the deadline. Another point, however, goes to ChatGPT for providing 8 different scholarships whereas Perplexity only provided 5. When asked for more scholarships like this, the chatbot easily came up with double the original amount. Lastly, perplexity gave more general scholarships, meaning scholarships that were open to anyone with the specific qualifications I provided. However, ChatGPT gave a plethora of options only open to certain races or financial groups.
When prompting the College Hippo Scholarship AI Engine, they provided similar results to both engines. it had most of the positives we found in ChatGPT and perplexity yet included both the popular options for scholarships and the niche ones for specific colleges. After testing the other engines, it is also significant to know that College Hippo prioritizes taking information directly from the official college website rather than information on social media or second-hand websites. This is another option that perplexity offers that clearly states to provide reliable data from certified sources. Overall, all three engines provide basically the same information for a student seeking specific scholarships.
The greatest advantage generative AI offers here is the room for specificity. Most websites and high schools offer wonderful resources honed for the general population of students, with many of them being solely for certain socioeconomic groups or minorities. Through scholarship search engines, the more information the student provides about themselves the more specific help they can receive. When searching using Google, we found that by changing the prompt slightly it would just result in the same top results. The sponsored scholarships would be the first to come up, and then the popular sites, no matter how accurate the content is to the student’s prompt. And unlike AI, more words and longer prompts would confuse the engine since it’s trained to only focus on keywords. Yet using AI we found niche scholarships that didn’t even come up on the first page of Google’s results. Next time you try to find scholarships, start with AI. Introduce yourself by grade, school, major, and gender to see the miracles of generative AI.