These Unknown AI Tools Give PhD Students an Unfair Advantage (Before Everyone Else Finds Them)

New & Emerging AI Tools for Academic Research

Key Concepts:

• AI Hallucination: The tendency of Large Language Models (LLMs) to generate factually incorrect or nonsensical information.
• Reference Verification: The process of confirming the existence and authenticity of cited sources.
• Roll-to-Roll Processing: A manufacturing technique for producing flexible electronics.
• Reporting Guidelines: Standardized sets of instructions for reporting research findings in specific fields.
• Semantic Scholar: A free, AI-powered research engine for scientific literature.
• Synthesis (AI-driven): The automated process of combining information from multiple sources to create a coherent overview.

1. Sitely: Source Verification & Authenticity Checker

Sitely aims to address the problem of AI hallucination in academic research by verifying the existence of sources. Users can utilize two primary functions: finding sources based on a research query and verifying references. The “find sources” function locates relevant papers, guaranteeing their real existence. The “verify references” feature allows users to input citations (individually or in lists) to confirm their authenticity.

While the tool duplicates functionality found in more established platforms, its unique strength lies in its reference verification. However, the tool struggles with multi-line references, requiring manual formatting adjustments. Sitely employs AI detection algorithms and cross-referencing to generate an “authenticity score” for each reference. The speaker notes that even partial references are flagged, demonstrating its ability to identify incomplete or inaccurate citations.

2. Academate: PhD & Presentation Practice

Academate is designed to alleviate the anxiety associated with presenting research, particularly during PhD defenses or conference presentations. The tool allows users to upload their work (though a thesis upload initially failed in testing) and engage in a simulated question-and-answer session with a virtual committee.

The questions generated by Academate were described as “impressive,” and the platform provides feedback on the user’s responses. The speaker highlighted the value of practicing with potentially challenging questions in a low-stakes environment. A key drawback is the credit-based system, with costs of $97 for three sessions, raising concerns about reliability and value for money at this early stage.

3. Thesisit.ai: AI-Assisted Thesis Generation

Thesisit.ai offers a range of tools to assist with thesis writing, including topic generation, introduction drafting, literature fetching, annotation, literature review drafting, methodology drafting, and data analysis support. The speaker expressed concern about the potential for misuse, particularly with the “Instafy my thesis” feature, which aims to generate a complete thesis with a single click.

Currently, the purchase function for the “Instafy” feature is disabled, potentially by design. The tool offers a topic generator with a somewhat lengthy question process. A disclaimer emphasizes that the tool is intended to overcome writer’s block and structure arguments, not to replace critical thinking.

4. Libright.app: Academic Writing & Formatting Assistance

Libright.app provides a suite of tools for academic writing, including manuscript reformatting for specific journals, in-Word editing and citation management, and peer review assistance. The tool’s integration with Microsoft Word is a notable feature, offering a seamless writing experience.

Key functionalities include reformatting manuscripts to journal specifications, peer review capabilities, and the generation of reporting guidelines checklists. The speaker noted the potential value of the reformatting and peer review features, even in the tool’s current state. An affiliate program is available, but the speaker declined to participate to maintain objectivity.

5. Ascent: Literature Search & AI-Powered Synthesis

Ascent is a research platform offering features similar to Consensus, including the ability to search for papers, compare them, generate citations, and read papers. A unique feature is the ability to select specific fields of study, allowing for more targeted searches.

Ascent’s most promising feature is its AI-powered synthesis capability, which analyzes and synthesizes information from up to 30 papers. While the synthesis process was slow during the demonstration (taking over 30 minutes), the speaker expressed enthusiasm for the potential of this feature. The platform is currently free to use, making it an accessible resource for researchers.

Notable Quotes:

• “The one thing that is always annoying about using large language models is that they can hallucinate. I have tested them. They hallucinate a lot.” – Regarding the need for source verification tools.
• “This is a tool to help you overcome writer's blocks and structure your…arguments is not a substitute for your own critical thinking.” – A disclaimer from Thesisit.ai, emphasizing the importance of original thought.
• “Supporting these early stage things will mean that we will have tools that we like going forward.” – Highlighting the importance of providing feedback and support to emerging AI tools.

Logical Connections:

The video progresses from tools focused on foundational research tasks (source verification with Sitely) to those addressing specific challenges in the research process (presentation practice with Academate, thesis writing with Thesisit.ai, writing assistance with Libright.app) and finally to a comprehensive research platform (Ascent). This order reflects a logical workflow for academic research. The speaker consistently emphasizes the importance of providing feedback to these early-stage tools to facilitate their development and improvement.

Conclusion:

The video showcases a selection of emerging AI tools with the potential to significantly impact academic research. While these tools are not yet fully polished, they offer innovative functionalities that address common pain points for researchers. The speaker advocates for actively testing and providing feedback to these tools, emphasizing that supporting their development will ultimately lead to a more robust and effective AI ecosystem for academia. The key takeaway is that early adoption and constructive criticism are crucial for shaping the future of AI in research.