Online: Trending Now

The Generative AI (GenAI) tools we know as ChatGPT-4, Gemini-Advanced and their competition in 2024 are all getting many more features than the versions we used just a year ago.

Speaking at the World Government Summit earlier this month, OpenAI CEO Sam Altman said GPT-5 will be ‘smarter’ and perhaps faster and more multimodal than previous versions of GPT. Altman apologized for perhaps sounding flippant in saying it would be “smarter,” but the OpenAI leader went on to explain that if the coming version of AI is smarter across the board, it is a truly significant improvement. This is not merely adding capabilities here and there, it is, rather, a holistic improvement.

There is a continuum of growth in size, speed, features and results that we are witnessing rolling out day by day and week by week. The competition among those leaders in this field is relentlessly driving small incremental improvement in the products. Yet, in a larger sense, we are also seeing GenAI improving in broad measures, as Altman suggests. As educators, we might say a student becomes smarter in the “four C’s” of holistic student improvement:

• Critical Thinking: Encouraging analytical and thoughtful decision-making.
• Communication: Developing effective interpersonal and expressive skills.
• Collaboration: Fostering teamwork, empathy, and cooperation.
• Creativity: Cultivating innovation and problem-solving abilities.

We see similar learning improvements when comparing undergraduate students to graduate students. Unfortunately, we in higher education, the faculty and administrators, are not getting smarter and more confident about GenAI at the same rapid rate. 

A recent survey was conducted by Cengage Group and Bay View Analytics to better understand attitudes and concerns of higher education instructors and leadership.  Dr. Jeff Seaman, lead researcher and director of Bay View Analytics, reported that only 16% of faculty and 11% of administrators felt prepared for the changes that GenAI will bring about in higher education. Seaman notes, “The delta between the expected impact of GenAI in higher education and the current ability of these organizations to adapt to this new technology is significant.”

It is critically important that we ensure our faculty and staff are prepared for the advent of more sophisticated AI tools as we move forward, because it is clear that advances are not slowing down. Truly revolutionary improvements will be upon us sooner than many of us may think. Those institutions that are prepared to take full advantage of the expanded abilities will surge ahead of their competition in efficiency, effectiveness and student outcomes, especially in preparation for the workplace where AI skills are increasingly valued.

For example, some faculty continue to rely upon ineffective tools to determine if assignments that are submitted have been written by GenAI. This is what I consider to be a mis-directed approach to serving students. GenAI applications are now fully capable of emulating the style and tone of the person prompting them. For example, here’s a recently posted routine that can effectively instruct a GenAI app to emulate your, or someone else’s, writing. This makes author detection very difficult if not impossible. Given the rate of computer-based tool adoption in business and industry, the distinction seems less relevant. If GenAI is considered as important of desktop tool as Excel, Word, Google Search and other common tools, the expectation to embed the product of these tools within daily work will be assumed.

As GenAI is more fully implemented in classrooms and workplaces, we are seeing new tools supported to enable Chatbots to enter into work groups and teams. GPT-4 recently enabled the capability to add an additional GPT to a conversation. So, one may create a three-way conversation between oneself, a GPT and an additional GPT. This opens the possibility of discussions between chatbots. It more fully reflects the environment that is evolving in businesses and industry where GenAI views are afforded the respect and consideration that human staff is given. 

Another recent advancement is the development of text-to-video capability. OpenAI earlier this month announced “Sora” which is a feature of GPT-4 that will generate stunning videos from detailed prompt text descriptions. In recent months, we have seen the advent of a host of Zoom AI note-taker bots. These common tools can often provide complete transcripts of a Zoom call, or an organized, annotated summary of key actions within the call. These are useful for a variety of reasons, including providing formal minutes of the meeting. Meanwhile, Gemini just announced that the 1.5 Pro edition can analyze massive files in a variety of formats including video. “Gemini 1.5 Pro can take in ~700,000 words, or ~30,000 lines of code — 35x the amount Gemini 1.0 Pro can handle. And — the model being multimodal — it’s not limited to text. Gemini 1.5 Pro can ingest up to 11 hours of audio or an hour of video in a variety of different languages.”

While the various versions of GenAI differ significantly, Professor Ethan Mollick of the University of Pennsylvania reports that although the leading two apps have different strengths, Google Gemini Advanced has significantly improved to the point of approaching a level of parity with the leader, ChatGPT-4:

“GPT-4 is much more sophisticated about using code and accomplishes a number of hard verbal tasks better – it writes a better sestina and passes the Apple Test. Gemini is better at explanations and does a great job integrating images and search. Both are weird and inconsistent and hallucinate more than you would like. I find myself using both Gemini Advanced and GPT-4, depending on circumstances.”

It is expected that GPT-5 will move OpenAI significantly closer to achieving Artificial General Intelligence (AGI). AutoGPT by Mindstream commented on the much-anticipated release of GPT-5 by OpenAI, noting in a report earlier this month: “While ChatGPT 5 is expected to showcase significant advancements in natural language understanding and contextual conversation capabilities, it is still uncertain whether it will achieve true AGI. Most AI experts estimate that AGI is still years away, and current AI systems, including ChatGPT 5, are likely to remain focused on specific tasks and domains. However, the progress and advancements made by OpenAI’s GPT models indicate that the pursuit of AGI remains a priority and will continue to shape the AI landscape in the future.”

If we are rapidly approaching AGI, one wonders how long it will be before we see the next generation of this technology, Artificial Super Intelligence (ASI)? If you are unfamiliar with the term ASI, USC Libraries defines it: “Artificial superintelligence (ASI) refers to a hypothetical form of AI that surpasses human intelligence across all fields, from creative arts to scientific research. Unlike contemporary AI, which excels in specific tasks, ASI would be capable of outperforming the best human minds in every domain.” And what will ASI mean for the future of higher education?

 

This article was originally published in Inside Higher Ed’s Transforming Teaching & Learning blog.