# Using AI in Literature Review (a rationalization)

By [PlaceCred](https://paragraph.com/@placecred) · 2024-05-19

---

While there are no prohibitions on using AI (that I’m aware of) in the [grant](https://explorer.gitcoin.co/#/round/42161/0x0f0b9d9f72c1660905c57864e79ceb409ada0c9e/0x0f0b9d9f72c1660905c57864e79ceb409ada0c9e-12) currently funding my work, I wanted to be transparent about my use of ChatGPT in the literature survey deliverable. While I can’t afford not to use AI at this point (more on that below), happy to respond to any stakeholders’ concerns this may raise and address them if possible.

**Explanation/Excuse**

Upon starting the literature survey, I quickly realized that what I’d promised in the grant, without using AI, would take a lot longer than originally estimated... In particular, a literature survey covering all of the relevant domains could easily take several months. On top of other deliverables.

I ended up getting roughly [$3,705 in funding](https://explorer.gitcoin.co/#/round/42161/0x0f0b9d9f72c1660905c57864e79ceb409ada0c9e/0x0f0b9d9f72c1660905c57864e79ceb409ada0c9e-12) from the [TEC’s Gitcoin Grant round](https://explorer.gitcoin.co/#/round/42161/0x0f0b9d9f72c1660905c57864e79ceb409ada0c9e). And I’ve spent ~31 hrs so far (breakdown below). Currently that’s ~$119/hr (yay!). But guestimating total hours without using AI, I’m looking at below minimum wage. Which is not financially sustainable for me…

![Hours breakdown (so far)](https://storage.googleapis.com/papyrus_images/d064f4fa0d5b72c3f8285fb3652e517fba3b0da379e1a590b2fa41d2ba939f76.png)

Hours breakdown (so far)

So, am using ChatGPT for parts of the literature survey. Unfortunately, that means sources aren’t cited :( I believe I’m still technically meeting the deliverable requirements, and delivering good value, for the following reasons:

*   ChatGPT is probably doing higher quality work than I could have, at least in summarizing topics generally. I’m also leveraging several years of relevant experience (including reading a large number of articles) in composing prompts, filtering/structuring material, etc.
    
*   For each section, I’m augmenting general summaries with knowledge and insights gained from ~5 yrs experience working on relevant projects. Namely [SourceCred](https://sourcecred.io/), a valuation/reputation protocol utilizing the PageRank algorithm, and [BridgeSpace Commons](https://twitter.com/BridgeSpacePdx), a physical community space in Portland I volunteered at most of last year that experimented with decentralized governance. I believe this ‘alpha’ should be generally valuable to those working on place-based reputation
    

I’m working on the literature survey now, and hoping to have it finished in the next 2-3 weeks. If you have thoughts or opinions on this, please reach out! I can’t work below minimum wage, but may be able to direct my efforts towards stakeholders’ needs.

---

*Originally published on [PlaceCred](https://paragraph.com/@placecred/using-ai-in-literature-review-a-rationalization)*