ˈper·iˌcule

Simon Willison’s 2024 retrospective on AI is loaded with insights. It’s worth reading the whole thing, but here are my top three takeaways: Despite being the most popular, GPT-4 isn’t the most advanced LLM—it’s not even in the top 50. I’ll admit that the few times I’ve used an LLM, I’ve defaulted to GPT because it’s just what comes to mind. Since reading this article, I’ve given Anthropic Claude a try, and I’ll have more to say about it in a later post. The point is, a heck of a lot of models are out there now, and it’s worth your time to think about which one is best to use, and that might depend on the use case at hand. ...

When we security types talk about “attack surface” or “threat surface,” we mean the part of our technology environment that’s potentially vulnerable. Think of your skin. We wear clothing to protect it from the sun’s harmful rays. A construction worker wears heavy leather to guard against wounds from tools and sharp materials. A cook wears an oven mitt to avoid burns. Any skin we leave exposed is vulnerable to these threats. We often deliberately accept these risks—for example, wearing just a swimsuit to the beach—sacrificing some security for some enjoyment and versatility. Ideally, we use good judgment and make a sensible, balanced risk decision. We can also use compensating safeguards (e.g., sunscreen) to limit the downside risk of our decision. ...