GPT-5 Is Here – Why I’m Excited, and Still Watching Closely

Posted on by Riko Hamblin
On August 7th 2025 I sat through OpenAI full GPT-5 launch. I still have some concerns but this is a step in the right direction. I remain optimistic on Generative AI as a technology with OpenAI as a major player in the market.

What I’m most excited about:
✅ The promise of “PhD level expertise in your pocket”, in a way reasoning as a service and the promise of the model thinking “the perfect amount to give the perfect answer”
✅ All prior models will be deprecated. This signals to me that OpenAI has very high confidence in this model as it pivots their entire user base to this new model.
✅ Improved safety – There is more nuance on inferring intent and giving safe responses without judging the prompt too harshly on maliciousness. Prior models would simply not respond if it is determined that the prompt is malicious. This new model promises to safely complete the response in dual use cases
✅ Increased context window and lower cost per million tokens – more capabilities for less money is always a plus!Where I will continue to monitor:
Hallucinations – when a model give false information as if it were true
Deceptions – when a model misrepresents the steps it took to get to an outputNew evaluations were determined to better gauge these dimensions and lower them. However, low is still not zero so I will continue to advocate users to vet the information they are receiving from these systems especially for highly sensitive areas like the intersection the intersection of AI with the health sector.I love the idea of putting knowledge into the hands of people and I strongly believe technology should enrich our lives however with the risk of hallucinations and deceptions still existing I remain cautiously optimistic on the health sector for example.All in all, as a Senior Software Engineer and AI Instructor by trade and builder at heart, I’m looking forward to building with this!