By Daniel Otunge
For starters, Artificial Intelligence (AI) is a collection of emerging technologies that leverage machine learning, data processing, and algorithmic systems to perform tasks that hitherto could only be done by humans. AI covers a range of skills, including automated reasoning, decision-making, and language processing.
For instance, Generative AI (Gen AI) can be used to create, edit, modify, or analyze content, including text, photos, videos, and sounds, as well as software code. There is an ongoing, silent yet highly active race among tech giants in the USA, China, and the European Union to produce the most innovative and most eco-efficient AI models. Thus, with AI models improving at a rapid pace, new Gen AI tools are being rolled out into the market for content creators, businesses, and states to use at almost a weekly cadence.
The latest in the market is OpenAI’s ChatGPT-5. It is an improvement on CPT-3 and GPT-4. With a new, simplified unified reasoning architecture, GPT-5 can determine in real-time whether your question requires a quick response or a more thorough evaluation. The result provides users with both speed and depth, eliminating the need to fiddle with complex AI settings and guess which model to use. It’s the closest yet to having a subject-matter expert, such as a lawyer, financial analyst, or doctor, in your pocket, says gurus like Dr Jeff Dalton, Chancellor’s Fellow in the School of Informatics, University of Edinburgh.
It is a leap for software development, one that is not just about writing code, but about shipping working software. Ask it for anything, and it will break the job into steps: write and test the back-end, choose colors and layouts, and hand back something that looks ready for users. Tasks that filled a developer’s week with GPT-4 can now be sketched in an afternoon. Think of it as a senior pair programmer who handles autonomy, collaboration, context, and testing all in one, Dalton says.
“The most significant development is that OpenAI is looking to reduce the number of models available to just the new GPT-5 suite. This means that those on premium subscriptions will not be required to select a specific model for a particular purpose. In essence, GPT-5 will choose the optimal model for them,” says Dr Junade Ali, Software Engineer and Computer Scientist, the Institution of Engineering and Technology.
Ali says OpenAI’s new model has the potential to improve user experience and reduce energy waste from powerful models being used unnecessarily. “This highlights the importance of society adopting environmentally friendly energy generation and storage technologies to address the demands of an ever-more technological society.” Nevertheless, the environmental cost of the energy usage of Large Language Models (LLMs) like GPT-5, Gemini, and Claude remains a key concern, as tech giants continue to explore nuclear power as a source of carbon-free electricity.
According to Prof. Anthony Cohn, Professor of Automated Reasoning at the University of Leeds and The Alan Turing Institute, who monitored the livestream of the launch of Chat-GPT-5, it appears to have improved performance and better functionality than its predecessors. For example, “the ability to generate software for a complete website, as demonstrated in the live stream, is impressive if it works in general. Also, the claimed improvements in safety and reliability are welcome. But the well-known weakness of LLMs that they “hallucinate” has only been “mitigated” according to the presentation, and it’s hard to imagine that this hallucination problem will ever be solved completely, given the way that LLMs operate—statistically predicting future text based on their training data.”
When asked whether GPT-5 solves hallucinations or safety worries, Dalton responded: “Solve is too strong, but progress is real. In everyday chat, the error rate falls from about twenty percent to roughly five percent. In a tough healthcare test, the hardest questions dropped from ten wrong answers in a hundred to fewer than two. The model also says ‘I’m not sure’ more readily when it hits its limits, exactly the behavior you want. We still need to address many of the safety issues surrounding AI agents, and this doesn’t resolve that.”
Pundits say that OpenAI’s numbers show conversation-level hallucinations falling from over 20% to under 5%, and the model is about three times less likely to go along with things when a user is wrong. For example, on complex medical evaluations, the most challenging questions dropped from around ten wrong answers in a hundred to fewer than two.
The livestream demonstrated some leading-edge performance. Where GPT-4 makes mistakes on roughly one response in five, GPT-5 is more reliable and competent, with misses closer to one in twenty. It feels more like a specialist you can keep in your pocket because it uses a unified “reasoning” model that chooses in real-time whether a prompt needs a quick reply or more profound thought, says Dalton.
Its pricing on its API interface is slightly cheaper than some of its other recent models, in particular o3. Thus, GPT-5 continues the path forward, making AI more accessible, reliable, and valuable. OpenAI has the model available to all US government workers.
Bottom line is that “GPT-5 is a significant incremental upgrade, smaller than the GPT-3 to GPT-4 leap, but big enough in accuracy, domain depth, and software-building skill to change everyday workflows. You can now prototype an app, or check a niche technical point, with much more confidence that the answer will be right and provided in a form you can use, ” Dr Dalton concludes.
As AI models are rolled out in quick succession by tech giants, governments, especially in developing countries, are under pressure to promulgate regulatory frameworks that will nurture the industry forward, harness its potential, and also guard against its misuse that could violate human rights.
