Regarding Sol 6, most are missing the point

First, do I think the marketing for this model is a bad look? Yes. But, I don’t think people realize what’s happening in the AI space, especially on Reddit. This shit simply is not profitable. Most of the people using AI are on the free tier (outside of enterprises). And enterprises are starting to retract on spend because the value simply isn’t there. Almost nobody is able to justify their AI spend. And OpenAI cannot afford to lose the enterprise spend. Enterprises are turning to open-source models, doing anything to save money. So OpenAI needs to focus on getting costs down. As far as I’m concerned that’s all this model release was, a cost play. It doesn’t have to be as smart as Sol 5.6 if it’s half the cost and is close. Most enterprises are shifting away from vibe coding anyway so the smartest models aren’t really needed. Hell, my team’s been using Luna for 90% of our work because we are competent engineers and use the tools for what we think it’s best for, crushing through ideas until we land on the one we like. Those ideas are our own though. I think this is a smart move. They need to compete with open source on cost, and given they haven’t been focused on doing so I expect to see some intelligence slippage while they figure it out. Trust me, intelligence isn’t going to matter anymore if they run out of people willing to give them money. Your subs will disappear overnight.