The video critiques the recent release of GPT-4.5 by OpenAI, highlighting its disappointing performance and high costs, which have led to skepticism within the tech community. Despite its aim for more natural conversations, the model struggles with basic tasks and fails to meet expectations, raising concerns about OpenAI’s future in the competitive AI landscape.
The video discusses the recent release of GPT-4.5 by OpenAI, which has been met with disappointment and skepticism from the tech community. Despite being the most expensive AI model ever produced, GPT-4.5 fails to impress with its performance, not achieving significant benchmarks or offering innovative capabilities. The model’s primary selling point appears to be its ability to engage in more natural, human-like conversations, but many feel this is insufficient to sustain the current AI hype. The presenter notes that the launch event was notably lackluster, with OpenAI CEO Sam Altman absent and interns handling the demonstration.
One of the key criticisms of GPT-4.5 is its high cost, priced at $75 per million input tokens and $150 per million output tokens, making it significantly more expensive than competitors like Claude. Access to the model is limited to Pro users who pay $200 per month. While the presenter acknowledges that the model emits “Chill Vibes,” this subjective quality does not compensate for its shortcomings. OpenAI introduced a new “Vibes Benchmark” to assess creative thinking, but the effectiveness of this benchmark remains questionable.
During a demo, the presenter highlights that while GPT-4.5 has a lower hallucination rate, it still makes numerous errors and lacks self-awareness. The model struggles with basic factual questions and does not perform well in programming tasks compared to previous models like GPT-3.5. The presenter expresses disappointment in the model’s performance on coding benchmarks, noting that it is not only less effective than its predecessors but also significantly more expensive to use.
The video also touches on the competitive landscape of AI models, with xAI’s Gro currently being regarded as the best model according to market sentiment. Although OpenAI is still favored to have the best model by the end of 2025, its odds are declining, raising concerns about its future viability as it transitions to a for-profit model. The presenter speculates that OpenAI may have struggled to make significant improvements in GPT-5, despite increasing the model’s parameters and computational power.
In conclusion, the video reflects a sense of disillusionment with the current state of AI development, contrasting expectations of rapid advancements with the reality of stagnation. The presenter emphasizes that while AI coding tools remain valuable, they are most beneficial to skilled programmers rather than replacing them. The video ends with a promotion for an educational platform, Brilliant, which offers resources for learning programming and understanding the underlying technology of AI models.