Seriously, everyone and anyone talks about LLMs these days like they have been doing it for ages. Everyone is posting online, having opinions and sharing their "expertise". How did people become expert in 6 months?<p>I am really curious about this one. Any clue?<p>There aren't any significant achievements. The ideas are more or less the same from 2017/2018 only scaled up. People have only recently started noticing the behavior like in-context learning. But some write with confidence as if they are the guru? Where does this confidence come from?
> How did people become expert in 6 months? I am really curious about this one. Any clue?<p>It is because they have something to sell you.<p>> There aren't any significant achievements. The ideas are more or less the same from 2017/2018 only scaled up.<p>This exactly.<p>It is by the grifter class who need to push the snake-oil as fast as they can. They are not building improvements on top of the paper, but are just throwing more compute power, tweaking parameters and most are just prompting their way to sell you something.<p>This is why you have a flurry of so-called 'experts' hyping about LLMs as if they are a new thing, but the hype came from the way it was packaged, like ChatGPT.<p>> But some write with confidence as if they are the guru? Where does this confidence come from?<p>Hype and over-promising to unsuspecting users who don't care about how it works or what the limitations or risks are.<p>These fake gurus and experts pontificating about LLMs are only doing it to kickstart their grift to VCs and sell their snake oil to the desperate. For example, Neeva.ai (founded by ex-Googlers) wanted to beat Google in the search engine race with LLMs but failed, despite being heavily VC funded.