Two years ago, on November 30th, ChatGPT burst onto the scene, igniting a global fascination with generative AI and transforming it into a must-have innovation for both consumers and technology professionals. Since then, ChatGPT has grown and the wheels of AI regulation have started turning.

TechRepublic asked tech experts how their work with ChatGPT has evolved both personally and within the broader tech industry.

New features introduced in 2024

Over the last year, OpenAI has:

• ChatGPT has expanded to include new formats such as ChatGPT Search and Canvas, the latter of which is partly designed to work alongside a coding application.
• Introducing the new flagship models GPT-4o and OpenAI o1.
• Partnering with Apple to support some features of Apple’s built-in AI.
• Announced: ChatGPT will remember previous conversations.
• Release of ChatGPT Search, marking OpenAI’s attempt to replace Google Search as the de facto portal to the rest of the internet.
• In October, Enhanced Voice Mode was introduced to select users, allowing users to speak out loud to the AI.

On October 3, OpenAI launched Canvas, marking a major experiment in the use of ChatGPT.

“To make AI more useful and accessible, we need to rethink how we interact with it,” the OpenAI team wrote in October when Canvas was announced. “Canvas is a new approach and the first major update to ChatGPT’s visual interface since our launch two years ago.”

How ChatGPT has improved in 2024

Graham Glass, CEO of AI course development platform Cypher Learning, noted that ChatGPT offers access to more sophisticated models now than in 2023.

“First of all, ChatGPT is getting better and better,” he said in an interview with TechRepublic. “And it has become more sophisticated, which opens up additional possibilities for using this technology.”

For the last year, Glass has used ChatGPT to develop software designs and software architectures. When he questions the technology about best practices or design trade-offs, he gets “the corpus of all the designs that everyone has ever done on that particular topic,” he said.

“It’s gotten smarter,” added Curt Raffi, chief product officer at Acrolinx, a company that uses AI to review content for technical documents and other writing-intensive tasks. He pointed out the improved performance of GPT-4o as well as OpenAI o1.

Raffi also explained that people have become more comfortable using ChatGPT. He works with engineers who have managed to prompt ChatGPT in a way that expresses specific business logic.

SEE: The first meeting of the International Network of AI Safety Institutes, held this week, aims to address the risks of advanced AI.

Glass likes that ChatGPT search provides up-to-date information and calls it a time saver on tasks like product comparisons. He also uses the enhanced voice mode to chat out loud with the AI.

Overall, ChatGPT’s additions over the last year have provided more options for people looking to use generative AI for technical work.

“The most significant way that generative AI assistants have changed programming and development in the last year is by giving people at different programming levels the opportunity to participate in software development to deliver solutions to real-world problems.” said Houbing Herbert Song, an associate at the Institute of Electrical and Electronics Engineers, said in an email to TechRepublic.

What can’t ChatGPT do in 2024?

AI is not immune to errors. For Glass, coding with ChatGPT often involves back-and-forth dialogue, including the AI ​​“remembering” details it may have missed.

“Although I think it is much more reliable in design (in 2024), it still makes a lot of coding errors,” Glass said.

For example, Glass said of a recent task that ChatGPT required 10 prompts to correctly build a function in JavaScript. This still saved him time, but it shows that ChatGPT is still limited. He attributed this in part to the fact that ChatGPT was trained on a finite, albeit huge, corpus of code.

Filev pointed out that ChatGPT has become so reliable that people may not easily notice when it makes mistakes.

“It’s getting so good that I started to let my guard down and I don’t know if that’s a good thing or a bad thing,” Filev said.

For many tasks, he started looking for hard sources on Google Search or Perplexity AI before using ChatGPT. These may be better places to find reliable sources, he said, while ChatGPT is better for brainstorming.

Regulations could impact ChatGPT

The last year has also revealed the limitations and possible regulation of ChatGPT. Raffi said his team is careful with AI-generated code following a lawsuit between developers and GitHub Copilot. The developers claimed that GitHub Copilot violated intellectual property rights by using open source code.

Raffi noted that commercial use of such code in marketplaces remains somewhat uncertain, making coding AI applications a cautious and exploratory process.

“Our intellectual property is embedded in our code, and if we were suddenly open or exposed to lawsuits, we could undermine the value of our company,” Raffi said.

How ChatGPT has impacted young developers

Another important development last year was the impact of ChatGPT on young developers.

“As this significantly improves developer efficiency and allows them to focus on higher-order design and innovation, the role of a developer – perhaps more importantly – is drastically shifting from creator to supervisor of AI-generated code,” said Dheerendra Panwar, senior -Member of the IEEE, in an email to TechRepublic. “This brings us to a very important question: Are we slowing down the art of coding?”

In some cases, junior developers are not hired at all because some tasks normally assigned to them are now being taken over by AI.

“These changes appear to be beneficial for experienced programmers by expanding their role and importance,” wrote Jen Stave, executive director of the Digital Data Design Institute at Harvard, in an email. “Because young developers often lack the expertise to detect problems such as AI hallucinations or inaccurate output, this incredibly important role falls to experienced programmers, who must now expand their responsibilities to mitigate risks such as AI-induced code errors.”

In other cases, junior developers may be better at prompt engineering than seniors.

“For young programmers, the story is more complex,” Stave wrote. “Generative AI reduces your dependence on collaborative problem solving and promotes more autonomous work. While this independence can increase productivity, it may not be a good thing for people who tend to thrive on human interaction and collaboration.”

Andrew Filev, founder and CEO of AI software development tool startup Zencoder, explained that typing ChatGPT might seem like a special skill. However, it reminded him that using Google Search was once a skill listed on resumes. Perhaps 2024 was the year ChatGPT began to influence the way tech professionals think about the portal to the rest of the internet.

“It’s becoming more and more an integral part of my day,” Filev said of ChatGPT. “It does give me a boost in productivity, but it doesn’t define me one way or the other, does it?”

ChatGPT Search and Canvas offer new form factors

OpenAI’s efforts to become a new portal to the rest of the Internet can be seen most clearly in ChatGPT Search and Canvas.

Raffi said ChatGPT search didn’t hold up very well to Google search because it lacked context and returned “pretty poor results.” However, he uses Canvas a lot.

“It changes the way we think about AI and ChatGPT,” he said. “It introduces an application layer and lets you think about the AI ​​APIs as the backend and more than the business logic behind it. It abstracts much of the confusing complexity of many editors.”

Because Canvas stores memories, it can reference previous changes to the code. Raffi called it a mix of application layer, backend and business logic layer.

There will be many changes in the coming years

2024 has shown that AI can’t do everything and that the number of transformative use cases may be slowing down. On the other hand, AI companies are training models to process more and more data, including improving the models underlying ChatGPT. The way professionals interact with ChatGPT has changed since 2023 and will likely look different in a year.

“Yes, there will be changes,” Filev said. He compared the rise of AI to the shift from punch cards to software programming. “But I think developers are used to change.”

“Technology is advancing and we are keeping up with it, and I think we can do a lot more and better as a result,” he added. “And ChatGPT is one of the good examples of technologies that help us.”