The year 2023 has witnessed a surge of interest in AI, with companies embracing the power of AI tools to enhance productivity. At the forefront of this movement is ChatGPT, a popular chat-based AI tool that enables human-like conversations, and its latest model GPT-5. OpenAI’s GPT-3.5, the underlying language model, has sparked discussions about the potential of AI while raising concerns about its impact on jobs and long-term implications. The main concern is that with GPT-5, as computational requirements and the proficiency of the chatbot increase, we may also see an increase in pricing.
Release date and delays
Following the release of GPT-3 in June 2020, OpenAI launched subsequent versions, including GPT-3.5 (davinci-003) in November 2022 and GPT-4 in March 2023. Although the release of GPT-5 is highly anticipated, OpenAI’s CEO, Sam Altman, has stated that the company is not actively working on GPT-5 and has emphasized the importance of addressing safety issues in their ongoing development.
GPT-4 introduced significant improvements over its predecessor, allowing input of images and supporting third-party applications through plugins. With GPT-5, users can expect even greater advancements, including support for larger queries and inputs. However, these improvements may come with an increase in pricing, as computational requirements and chatbot proficiency continue to evolve.
GPT-5 aims to address concerns regarding factual accuracy and hallucinations in AI-generated responses. While GPT-4 showcased improvements in these areas by being trained with more recent and diverse datasets, GPT-5 is expected to further enhance fact-checking capabilities. OpenAI acknowledges the importance of eliminating incorrect responses, especially in critical fields like medicine and education.
Multimodal capabilities
GPT-4 has enabled limited web search functionality through a dedicated web-browsing plugin, allowing access to up-to-date information. However, the challenge lies in filtering reliable results from potentially unreliable sources. OpenAI’s future models, such as GPT-5, may focus on refining these capabilities. Additionally, GPT-5 could expand its range of data processing to include audio, video, and other forms of multimedia, making it more versatile and applicable across various domains.
Is artificial general intelligence (AGI) on the roadmap?
While AGI remains a theoretical concept, its association with GPT-5 has sparked speculation and debate. AGI represents the stage where machines possess self-aware consciousness and autonomous learning abilities similar to humans. OpenAI has released tools like Auto-GPT, which provide a glimpse into the future potential of AGI. However, concerns regarding the risks associated with AGI, as highlighted by OpenAI’s CEO, Sam Altman, call for a global conversation on governing, distributing benefits, and ensuring responsible access to AI technologies.
OpenAI’s GPT-5 is eagerly awaited because of the potential advancements it may bring to language models. Expectations include increased power, improved factual accuracy, multimodal capabilities, and progress toward AGI. However, OpenAI’s emphasis on addressing safety concerns and the ethical use of AI may delay the release of GPT-5. As AI technologies continue to evolve, balancing innovation and responsible development will be crucial for ensuring the beneficial integration of AI into various industries.
Our take on the AGI concern: While it may be an exaggeration to expect GPT-5 to conceive AGI, especially in the next few years, the possibility cannot be completely ruled out. Keep watch, people!
