Raw LLM Responses

Not familiar with “Bold”, but familiar with the Gartner hype cycle. It’s anyone’s guess when we’ll enter the trough of disillusionment, but surely it can’t be *that* far off? I’m uncertain because right now, there’s such a massive amount of financial interest in propping up LLMs to the breaking point, inventing problems to enable a solution that was never needed, etc. Another challenge is since LLMs are so useful on an individual level, you’ll continue to have legions of executives who equate their weekend conversations with GPT to replacing their entire underwriting department. I think the biggest levers are: 1) enough executives get tired of useless solutions, hallucinations, bad code, and no ROI 2) the Altman’s of the world will have to concede that AGI via LLMs was a pipe dream and then the conversation will shift to “world understanding” (you can already see this in some circles, look at Yan LeCun) 3) LLM fatigue - people are (slowly) starting to detest the deluge of AI slop, the sycophancy, and the hallucinations - particularly the portion of Gen Z that is plugged in to the whole zeitgeist 4) VC funding dries up and LLMs become prohibitively expensive (the financials of this shit have never made sense to me tbh)