TrevorJ

Not sure if the format for this article is standard these days, but oh man do I hate it.

Coatings that don't cut down on reflections is the biggest issue I've had the various times I have used this route.

there are third party vector databases though

100% agree. This guy is next level.

"The last few moments have disproven this idea that ships need to stop hitting icebergs. Now, dear guests on this great ship Titanic, let us go back to enjoying our dinners!"

Timeframe matters a lot when you run an experiment.

"a recession is when your neighbor loses his job; a depression is when you lose yours"

There are some laptop desk arms that you could use to extend them out to either side. If you search on amazon you can find a bunch. I use the one that is only articulated in one direction (not the fancy ones). Works great.

This assumes that categorizing a hot dog as a sandwich is logical though.

Hard disagree, as someone in the industry.

Mental health conditions are typically defined by a collection of symptoms though, aren't they? I am not suggesting there is no underlying cause, but our ability to detect and quantify that cause is lacking, so defining disorders based on a collection of symptoms is what we are mostly left with in many cases.

Sure it does, if those tokens appear in the training data.

Businesses will charge what the market will bear. And if 'the market' can get government loans to pay for tuition, then the market will bear an awful lot.

I agree that most people are not biased towards action, and fixing that is the first step. However, the article is clearly aimed at people who have made that change and now are faced with the more subtle question of "how do I think about the problem such that the action I choose is likely to be appropriate."

The issue here is that running a website at all is hitching yourself to the google bus.

  Location:Colorado
  Remote: Yes
  Willing to relocate: Yes
  Technologies: Python, game engines (Unreal, Unity et.al.) DCC's (Max, Maya, Blender, Adobe etc), AI assisted Artistry, Git, Linux
  Résumé/CV: Available on request
  Email: TJ345675643@protonmail.com

I have 20 years of experience in my field, and have enjoyed working on a huge variety of projects from DoD to Fortune 500 to small mom and pop shops.

To build on this, personal cybersecurity is complex in general, and only seems 'simple' if you've been blessed to grow up around computers, the internet AND have more than a simply utilitarian interest in technology.

Increasingly, the world assumes that all three of the above are true for everyone, and anyone else gets marginalized.

It's a very underrated side effect of this whole LLM thing: We've created a super compact representation of human knowledge in a form that requires a FAR less complex tech stack to get the information 'out' of in the future.

A year ago, a lot of this information only existed on the internet, and would have been nearly impossible to recover in any cohesive unfragmented form if the lights were to ever go out on our civilization.

Now the problem space has moved simply to "find a single solitary PC that will still boot up", and boom, you have access to everything.

I think we just created our Rosetta stone.

This seems fraudulent, given the fact that the AI is creating scenarios that would not actually work in the real property. Books on a windowsill that is way too narrow for books, furniture drawn too small, making the room look huge. Seems like a lawsuit waiting to happen.

It's not a replacement for search, it simply makes search worse.

I strenuously disagree with this. Search has become all but useless in the last few years. LLM's provide a huge jumpstart on understanding, specifically, what to search on to find the canonical data you are after.

The reason some of these vehicles ARE so big IS the regulations. We could look at rolling back THOSE changes first.

this elides the discovery problem somewhat.

This is not very much different than the current state of the indie market.

As someone who's learning was significantly accelerated by the "written tutorial" phase of the internet this would be a really great little tool. I find video tutorials to be far more cumbersome than text+ images.

If I had to guess at what the next leap for AI would be, it looks something like a collection of small and large special-purpose models, with a model on top that excels at understanding which sub-models (if any) should be applied to the current problem space.

interesting, thank you.

I think with RAG it's pretty reasonable. Put your corpus in pinecone or some other vector store and relevant sections get injected along with your prompt which lessens the burden on context window.

That's pretty cool - let us know if you do!

That's a pretty slick plugin I will have to check it out, thanks for the suggestion.

Yeah, it's a good point. Have a feeling we will see a lot of sneaky security issues around LLM's over the next few years to say the least.

I'm pretty interested in this as well. I have moved from Notion to Obsidian for my personal notes, to-do lists and errata in preparation for this since obsidian uses local plaintext files.

What I would love to get working at some point is allowing an LLM access to my schedule, notes and goals and then have it help prompt me at appropriate times. "Hey, TJ I noticed you haven't worked out this week, it's sunny today this might be a good time". That sort of thing.

There seem to be good tooling around agents, prompt engineering, RAG etc. The 'glue' around getting the LLM to help figure out when appropriate time(s) to check in with me is the bit I am missing, but that's probably mostly down to me being an artist and only a very very JR hobbyist programmer though.