<p>It’s remarkable how often you need to be dramatically upgrading your AI architecture given the pace of progress in AI models right now. <br>

<br>

If you’re building agents, you basically need to throw away large parts of previous work that you setup to compensate for model limitations every few quarters. The systems you built to mitigate context window limits aren’t useful anymore, and for many use-cases it’s easier just to throw more compute at a problem today in ways that wouldn’t have worked previously.<br>

<br>

If you’re deploying agents in a workflow, you likely need to equally be rethinking your core systems at about that same frequency. The way you would deploy agents in an enterprise 18 months ago is entirely different from the best practices that you’d have today.<br>

<br>

This is partly why everyone’s working so hard right now. Right as a best practice is solidified, models improve dramatically, and that old work is rendered obsolete. Unclear that this lets up anytime soon, which is why the it pays to be so wired in right now.</p>

<hr/>

<blockquote>

<b>Sam Hogan 🇺🇸 (@samhogan)</b>

<p>

<p>most of tooling around llms was built for a world that largely doesn’t exist anymore<br>

<br>

RAG, GraphRAG, Multi Agent Orchestration, ReAct frameworks, prompt management/versioning tools, LLMOps tooling, eval tools, gateways, finetuning libs, etc<br>

<br>

all obsoleted in in the last 3 months</p>

</p>

<footer>

— <cite><a href="https://nitter.net/samhogan/status/2045359777270927699#m">https://nitter.net/samhogan/status/2045359777270927699#m</a>

</footer>

</blockquote>

为什么值得关注

能改变理解方式，而不只是重复常识；符合当前抓取需求；它提供了新的理解或解释，而不只是表面观点

来源：x，领域：tech，保留分：0.65