RSSAmplifier

Blog

insight forge

Use LLMs Better!

insfor.wordpress.comRSS feed ↗10 posts

Latest posts

Can AI be used safely?

There are risks to using AI. But there is also an enormous amount of pressure to use AI. The message is, USE IT FOR EVERYTHING OR YOU RE A LUDDITE!!!!! This is not a constructive message. Very few people want to slap AI into place without even giving a thought to whether it s appropriate or effective. [ ]

The Dawn of the LGTM (AI) Apocalypse

Badly reviewed code has consequences. I've seen it bring down a site the moment it was deployed. And I've seen it get bogged down in QA because of unexpected handling of error cases. And I've seen it cause issues for users that QA, somehow, didn't catch.

Let’s Talk Through Using AI

In July of 2025, I quit my programming job to focus on AI. Okay. To be honest, the focus on AI was only part of the reason. But it s the part that matters here. I use AI every day and for a lot of different things. I use it to write code (I m currently developing [ ]

A Nuisance a Day

Between these two things, we see a potential side effect of new technology that product development teams rarely, if ever, think of. What will law enforcement do? I've been involved in the development of quite a few (mostly web) applications and when I have raised questions along those lines ...

Unintended Side Effects

I've been thinking a lot about unintended side-effects of technology.

Announcing Big Changes

Consequently, I feel that it would be a conflict of interest for me to continue offering the consulting services I've been offering. So I'm going to stop doing that. If you were thinking of engaging Insight Forge to assist your AI efforts,

The Riddle of AI Regulation: A Rant

... there is so much wrong with this entire idea that I hardly know where to start. The first thing that comes to mind is that they are trying to regulate something they don't understand. This can't work.

The Trouble With LLM Benchmarks

So most benchmarks are only relative to themselves. LLM X performs better on some benchmark than the last version or than LLM Y. Good to know IF ...

Prompt Engineering: Final(ish) Thoughts

... Suppose they had been charging serious money for the image generation service, though? In that case, they would be facing potential lawsuits and loss of serious revenue. And why?

Prompt Engineering: Security

Security is a never ending problem for anything to do with computers and especially things to do with the Internet. When you've attempted to integrate an LLM into your system, there a re extra concerns that need to be considered.