Prashant

mostly deterministic

[Prashant @capeandcode]

got claude certified!🏅

[Prashant @capeandcode]

My experience in corporate is getting worse as my experience is increasing.

[Prashant @capeandcode]

What happens to facebook now?

[NSA Cyber @NSACyber]

Effective Immediately, all organizations using PHP for their systems are advised to be prepared for an update addressing two critical vulnerabilities. PHP is widely used for web applications. These vulnerabilities often provide a direct path to the server's file system or database. If an immediate update is not feasible, it is recommend that organizations take the system offline until it can be fully mitigated. NSA will follow up shortly with more specific guidance.

[Prashant @capeandcode]

@serious_mehta As a monk usually

[Prashant @capeandcode]

@AdiPolak nor for ppts.

[Prashant @capeandcode]

Summarize and put this on a billboard.

[Brad Traversy @traversymedia]

How can newbies learn how to use AI if they don't write code first? Nobody seems to talk about this. The only reason anyone understands any software architecture at all is because they went through the process of learning and writing code. We only seem to think about the developers of today using AI.

[Prashant @capeandcode]

@AnkitaxPriya @Heypocket Why not just record a memo and send to any llm? Genuinely asking.

[Prashant @capeandcode]

@FrancescoCiull4 What’s stack overflow?

[Prashant @capeandcode]

For any of technical solutioning and architectural brainstorming I always ask it to create interactive HTMLs.

Easy to understand and explain as well!

[Andrej Karpathy @karpathy]

We'll be spending a lot more time trying to understand the outputs of language models. A few thoughts, tips & tricks:

Writing. Something I've had success with: Ask your LLM to explain something in ASD-STE100, it's a controlled language specification originally developed for aerospace maintenance documentation. LLMs well-versed in this language and it comes with heavy constraints on clean writing style that I often find a lot more readable. Sometimes I've tried to soften it a bit e.g. ask for "80% of the way to ASD-STE100" because the spec is quite stringent. But even better:

Diagrams / images. Instead of writing, ask your LLM to create a diagram. These can be a lot easier to process, parse, and understand. But even better:

Web pages. Ask for output "in HTML" to get a beautiful, interactive webpage. LLMs are getting really good at frontend and can create beautiful experiences, animations, etc. But even better:

Explainer videos. The output format I am most bullish on is fully custom / bespoke explainer videos generated on any arbitrary topic. Experiment with things like "Create a 3b1b style video explainer on X. Use my ElevenLabs API key for audio narration". (you'd need an API key for the latter or you can ask your LLM to find you decent free alternatives that use your local compute). This is actually starting to work!

In summary:
- As LLMs get better, they will do more and more of the legwork autonomously, and a lot more of our work will rise up the abstractions into oversight and understanding.
- Luckily, LLMs can help here too because as intelligence and code are increasingly abundant, you can ask for large, custom, discardable software artifacts (e.g. web apps, video explainers) that would have never made sense to create before. Push the boundaries here and you'll be surprised.

[Prashant @capeandcode]

Maybe the AI race was never about models competing with each other.

Maybe it’s about us humans racing to keep up with them.

[Prashant @capeandcode]

JEV listened to me I guess.

[Prashant @capeandcode]

I still struggle to understand how to get a reliable confidence from an LLM.

Not a random number but a number that can be trusted.

[Prashant @capeandcode]

The amount of ppts you have to deal with is directly proportional to the number of promotions you are getting.

[Prashant @capeandcode]

@kokaneka Been seeing a lot of jev on the TL. First post to explain it somewhat.

[Prashant @capeandcode]

I still struggle to understand how to get a reliable confidence from an LLM.

Not a random number but a number that can be trusted.

[Prashant @capeandcode]

@SukhPunn Yeah. Correct

[Prashant @capeandcode]

Why is it that we cannot keep two files with identical names within the same directory?

Yet we can delete both and they live together in the bin!

[Prashant @capeandcode]

One of reasons I like iOS.

[Dhananjay_Tech @Dhananjay_Tech]

My biggest issue with iOS has always been missing Notification icons in status bar.

When the phone is unlocked, there is no way to know which app has new notifications without swiping down!!!

[Prashant @capeandcode]

This is applicable to a large population of humans as well.

[Tivadar Danka @TivadarDanka]

LLMs are statistical bullshit generators. They are trained to say what you want to hear. Not what you need to hear.

Yeah, I know, there’s reasoning, tool calling, memory, and other neat tricks. Still, fundamentally, generative models are stochastic parrots.

[Prashant @capeandcode]

@chrisalbon I love using that setup but mu sweaty palms ruin the experience. So switched to external kbs.

[Prashant @capeandcode]

@theskindoctor13 विनाश काले विपरीत बुद्धि

[Prashant @capeandcode]

All tax and no perks makes common man an angry boy.