There are currently a number of prominent Large Language Models, such as GBT-4 from Open AI, Gemini from Google, Claude from Anthropic, and Llama from Meta. There are also programs that are essentially wrappers built around one of these LLMs. Some of the products that are wrappers may be easier to use from a consumer perspective, but the educational mission of some schools, such as AU’s Kogod School of Business, to deeply train students in how to use AI will probably require exposure to the base-model rather than a consumer-oriented product.
The advancement and deployment of software applications is perhaps the clearest visible sign of AI’s rapid development. These are some of the best-known tools that have become available:
- The one that has so rapidly and strongly brought AI (back) to the center of attention, prompting the report you are currently reading: the “Chat Generative Pre-trained Transformer,” far better known by its short name ChatGPT, by OpenAI. In early 2023, it was “…estimated to have reached 100 million monthly active users in … just two months after launch, making it the fastest-growing consumer application in history…” (Hu, 2023)
- Gemini by Google, which is Google’s AI and claims to be: Your tutor, your data analyst, your creative partner, and your coding helper (Google, 2024).
- Bing Chat (now Copilot) by Microsoft, based on ChatGPT's GPT-4. Microsoft has increased its multi-billion dollar stake in ChatGPT developer, OpenAI, over several years (Microsoft and OpenAI Extend Partnership, 2023). Bing Chat was rated “as less helpful and trustworthy than ChatGPT or Bard” in a mid-2023 study with 18 participants (Liu et al., 2023).
Focused more directly on academic pursuits are the following AI tools (Jones, 2023):
- Elicit, which offers to “automate time-consuming research tasks like summarizing papers, extracting data, and synthesizing your findings.” It comes in a free version to “explore what Elicit has to offer” and two for-fee pricing models for individuals and for “teams, companies, and educational institutions of 50+ people.”
- Consensus, “a search engine that uses AI to find insights in research papers.” Like Elicit, it comes in a free version, a more capable for-fee premium version, and an enterprise version.
- TLDR This, which “helps you summarize any piece of text into concise, easy to digest content so you can free yourself from information overload.” The web site states that this tool is developed by a company called “Tridev” but provides no information about the company.
- AskYourPDF - “Ever wished your documents could talk? With AskYourPDF, they can! Powered by ChatGPT we breathe life into your documents, making them interactive and engaging.” Like other products, made available in free and multiple paid versions. The web site shows “BlockTechnology OÜ” as the developer but provides no information about the company.
- Julius, an “AI data analyst & more” for numeric data by Caesar Labs, Inc., “a small AI lab in San Francisco.” Most explanatory pages of the web site cannot be accessed without signing in; they “offer a 50% discount to students and other members of academia” but do not list pricing (again, without login).
- Powerdrill “is built for bridging your data and AI. It provides the services and platform for no-code and one-stop integration of your data and OpenAI large language models (LLMs) for intelligent Q&As and ecosystem interaction.” Again, there is a free and for-fee version. The company listed in the terms of service and the Privacy Notice is “ROCOSKY” but no information about it is provided; it may well be ROCOSKY TECHNOLOGY PTE. LTD. listed in the D&B Business Directory.
- Inciteful, with the goal “…to give the world free tools to help accelerate academic research. … Unlike a traditional search engine, citations are the cornerstone of all of our tools. Building these tools for all academic literature has only recently been possible with the rise of open scholarly bibliographic data … To date we have two tools. Paper Discovery and the Literature Connector. More tools are actively under development.” The focus here is on the connections between papers, built on the data and/or inspiration sources OpenAlex, Semantic Scholar, Crossref and OpenCitations. (Whether this qualifies as an “AI tool” is arguable; although listed in an article about such, it does not claim to be on its own web site.)
None of the inclusions of a tool in the above listing constitute a recommendation. Noteworthy in this sampling of tools is the wide range companies behind them, from the largest global players in computing and “search” to the most obscure, which relates to the question of what and how much information should or might users entrust to these tools?
The rapid development of software in the AI space is not only striking on a tool-by-tool basis, but in the emergence of a development ecosystem: “Those building or deploying AI applications now have multiple third-party infrastructure and platform services which expedite the process and reduce its complexity. A developer may build something end-to-end, but it is likely that they will assemble applications or models building on an increasing rich infrastructure and tool ecosystem. An AI stack is emerging, with capacities potentially provided by multiple players at each level.” (Generative AI and Libraries, 2023)
