Let's face it: after endless teasers on X and AI evangelists heralding the era of true AGI, GPT-5's arrival feels like a mixed bag that didn’t quite live up to the fireworks show everyone anticipated. It’s not the seismic leap to AGI that some predicted—no video inputs, no revolutionary multimodal features beyond what’s already available, and it’s rolling out with users scratching their heads. AI somehow feels more corporate—no more encouraging remarks, no more emojis. The new GPT sounds less like a friend and more like an ambitious work colleague.

What stands out: As the dust settles, a vocal minority of developers and researchers is already lobbying OpenAI for expanded access to GPT-5 Pro (cite tweet to Altman), not shying away from the rollout glitches but pushing to maximize its potential in production. These aren’t casual AI enthusiasts, but professionals recognizing value in highly expensive subscriptions. The question arises: if people see such value in these costly models, maybe OpenAI’s future isn’t $20 a month, but $2,000.

To me, the launch of GPT-5 reflects how AI systems are progressing. In that sense, we see divergent criteria for what is deemed “good” AI. We’re used to Apple’s presentations with clear metrics of improvement: faster chips, better screens, longer battery life. While benchmarks have traditionally been used to emulate this technical progress, it’s becoming clear that “progress” in large language models means something entirely different. GPT-5 may perform better on benchmarks than GPT-4o, but this is coupled with what users perceive as a downgrade in amicability. GPT-4o was nicer and softer but often agreed with users when it shouldn’t have. For writers, educators, and everyday creators, the loss of that "spark" has led to widespread complaints on Reddit and X, where over 50% of post-launch discussions skewed negative. https://x.com/ai_for_success/status/1956724160673275997?referrer=grok-com

Ultimately, the launch of GPT-5 signals one thing: with OpenAI’s mission to achieve dominance in business spheres and rapidly growing adoption in these sectors, they’re adapting their models for tasks like programming and finance, slowly shifting ChatGPT’s perception from a friend to a work colleague. ChatGPT now sounds more serious, straight to the point, and less empathetic.

The rollout itself has only accelerated these divides. While personal users were the key to OpenAI’s success, this launch suggests they might have different plans. OpenAI knows how quickly this market moves. When the programming world shifted to Anthropic and OpenAI scrambled to keep up, it might have realized the only way to remain relevant is through business. Consumer behavior can change rapidly, but business behavior is highly predictable. The saying “no one ever got fired for buying IBM” might soon become “no one ever got fired for buying OpenAI.” With drastic consumer shifts between Gemini, Anthropic, Grok, and others, securing business behavior—once achieved—would be highly beneficial. In that sense, as long as OpenAI remains relatively on par with competitors, its place in the AI space is secured.

Thus, GPT-5 might not be the breakthrough some users hoped for, but it may have secured OpenAI’s future and funding in the ever-changing AI landscape.