Are you a graduate student trying to navigate competing priorities, looming deadlines, and an ever-longer to-do list? Generative artificial intelligence (GAI) tools are becoming an increasingly helpful resource for students and researchers.

In last month's SPSPotlight feature, we shared the first part of an in-depth recap of GAI programs that can support your work. In the final part of this series, we explore the topics of AI-detection software and AI image creation, as well as several additional cutting-edge large language models that can supercharge your academic success. Let's dive in and see how each GAI tool can help you achieve your academic goals and research potential. If you've tried any GAI cutting-edge platforms, please let us know your thoughts or reviews.

AI-Detection Software & AI-Generated Images

Part 1 of our GAI guide discussed the importance of gaining academic approval from faculty members or advisors before using these new technologies. The technological rise of artificial intelligence tools now brings us additional platforms built to detect AI-generated writing. However, some new AI-detection tools like, GPTZero, or are still in beta development. In preparation for this article, I tried these new AI-detection tools and even chatted with the developers behind the GAI engines. Developers confirmed that any writing checked via Grammarly would rate high on AI-detection platforms as AI-generated text, even if that is not true.

If you are a student who regularly uses Grammarly, Microsoft Editor, or Google Workspace add-ons for grammar, plagiarism, or spelling - these new AI-detection tools consider the writing as AI-generated! It was surprising to learn this bug in the AI-detection tools, as most students use some sort of grammar or spelling tool. It's a good idea to let your professors know what grammar AI tool you are using.

I've been experimenting with GPT-4's DALL-E AI model developed by OpenAI, specifically designed to create images from text descriptions and generate content. DALL-E interprets the user's prompt to visualize concepts described in natural language to create novel images representing the user's description and theme provided. GPT-4's DALL-E produces visual representations uniquely showcasing the AI's ability to understand and merge different elements into AI-generated imagery. Note: This article's header image was created by DALL-E when prompted for visuals of generative artificial intelligence tools.

Emerging GAI Platforms

The landscape of GAI platforms for academic research and education is rich and varied, each offering specific strengths tailored to different needs. From enhancing literature reviews to providing detailed analyses of scientific citations, these platforms represent the cutting edge of AI's application in academia. Users should navigate AI's limitations and ethical considerations to leverage these technologies effectively and remain cognizant of GAI hallucinations or false responses. By understanding these platforms' unique features, benefits, limitations, and risks, researchers and students can make informed decisions that align with their specific requirements, thereby advancing the pursuit of knowledge and innovation.

1. Google's LaMDA

Google's experimental LaMDA focuses on generating contextually aware responses for more natural conversations. It is being designed for a wide range of topics, making it versatile for various applications.

  • Benefits: Designed for natural, wide-ranging conversations, LaMDA excels in generating context-aware responses across various topics.
  • Features: Contextual awareness, comprehensive topic coverage.
  • Limitations: As with any conversational AI, it may occasionally produce irrelevant or inaccurate responses.

2. Anthropic's Claude

Claude AI, Introduced by Anthropic, aims to generate informative and ethically sound responses. Its focus on safety and alignment makes it a compelling choice for prioritizing ethical considerations.

  • Benefits: Prioritizes ethical considerations and safety, making it a trusted choice for generating responsible content.
  • Features: Ethical response generation, safety, and alignment focus.
  • Limitations: The emphasis on safety may occasionally restrict creative or unconventional responses.
  • Costs: Try for free by signing up, with tiered paid plans available.

3. Meta AI's LLaMA

LLaMA stands out for its efficiency and scalability, yet as with most Meta developments, it requires some coding experience. It is designed to be accessible on lower-specification hardware, demonstrating Meta AI's commitment to democratizing AI research and development. However, LLaMA might not be the most user-friendly GAI tool available and is mainly for coding development.

  • Benefits: LLaMA's efficiency and scalability make it accessible on a wide range of hardware, democratizing AI research.
  • Features: Efficiency, scalability, accessible AI research.
  • Limitations: As a newer entrant, its ecosystem and support network are still developing.
  • Costs: There are paid tiers that depend on usage. However, LLaMA 2 claims to offer a free version.

4. Baidu's ERNIE

ERNIE incorporates knowledge graph technologies to enhance text understanding and generation, particularly in Chinese. It underscores the importance of specialized models for different languages.

  • Benefits: Specializes in understanding and generating Chinese text, incorporating knowledge graph technologies for nuanced text interaction.
  • Features: Knowledge graph integration, specialized in Chinese language.
  • Limitations: Its specialized focus may limit its utility for non-Chinese text applications.
  • Cost: Requires account, open source.

5. Microsoft's Turing-NLG

Turing-NLG, with billions of parameters, is Microsoft's effort to create powerful language models for natural language understanding and generation tasks. It marks a significant advancement in conversational AI technologies, yet it requires coding abilities and may not be as user-friendly as other GAI tools.

  • Benefits: With billions of parameters, Turing-NLG is at the forefront of conversational AI and designed for complex language understanding tasks that integrate with Microsoft Bing and other Microsoft products.
  • Features: High parameter count, advanced conversational capabilities.
  • Limitations: The size and complexity of the model may require substantial computational resources.

The possibilities with AI in academia are truly endless. Feel free to experiment and find the LLM that best suits your research style. With the right partner in GAI, you can conquer those deadlines, refine your research, and push your academic journey to new heights!

The SPSPotlight team is excited to hear about your personal experiences with GAI tools! Share your thoughts here.