JambalayaJimbo

What do you even consider advertising?

If I make let's say a children's toy, how am I supposed to let people know that this new type of toy even exists? That WHOLE process is advertising, no? Whether I give these toys away for free at a Santa Clause parade, pay department stores to put my toy in prominent shelves, or buy space on Cartoon Network.

How do you get even a single buyer without advertising?

How exactly can you ban an open source model? It’s effectively an algorithm. The US tried to do this with cryptography algorithms already

Even if the process of creating software was 100% cost-free and bug-free (which we know isn't true, even the most efficient LLMs actively burn tokens and cost you money) - pre-built software still has the advantage of being opinionated and having features I never thought to ask about.

I might have the general desire to play a video game with my friend where we shoot WW2 like guns together. I'm willing to bet that a video game company with artists, game designers, historians, etc. is better able to articulate that into existence than I can. Especially given I might not be willing to spend the time to do that...

If you don’t do chores are you able to perform magic? It’s an important question…

IMO there’s a ton of room for innovation in the IDE space.

It looks like software dev is going the way of natural language as the final spec that humans see. Regardless of whether this is a good idea or not, it’s happening. My boss asked us not to write any code manually starting in Jan, and now has hinted that we should stop reviewing code too.

Natural language as spec obviously has a ton of limitations. And here an opinionated IDE that can define this sort of workflow better is primed to gain a lot of market share, devs clearly aren’t fussed about switching between IDEs right now.

I should be able to, for example, click on a plantuml diagram and see what the endpoint definitions are, See where the request schemas are.

If I have Agents.md files scattered across my codebase I would like my IDE to help me manage them

Have you ever tried to use an LLM to add a feature to a really nice, pre 2023 codebase? It’s incredible how much easier it is to do, how much of a difference you instantly feel

We’re conditioned to interact with language models as chatbots and in that sense strong language understanding (implicit - some sort of world knowledge), is probably necessary for that?

But I’m sure we can have a small model that’s really strong at programming concepts, JavaScript syntax, and that’s about it. You’d interact with it differently, at specific seams in your code base - review a PR, merge two functions together, investigate these logs.

Or maybe I’m just not adequately absorbing the bitter lesson. Idk

Sure that’s a valid use case, 100%.

Program specific permissions (separate from the user operating them) are part of the Linux permissioning system already right? That doesn’t seem like an issue to me.

I guess the main problem would be finding an API client that can easily plug into your harness, with a nice UI for turning specific APIs on and off.

What do you mean by sandboxing bash? Isn’t this about just having a tool like curl or Postman?

Implanting an MCP client in your agent code isn’t all that different from calling requests or whatever

Measuring area is a different use case to looking at the world map to get a sense of relative sizes - of continents, countries, regions, etc.

This in turns influences how you think about history, population density, travel times, etc.

Re write the assistant message and see how easy it is to bypass a system prompt. How far do you have to stretch to get a human analogue? Short term amnesia?

Your example actually demonstrates why anthropomorphism is a bad idea.

LLMs are vulnerable to classes of attacks that humans just aren’t. In your framework, the way to prevent attacks is to… invent human consciousness?? It’s an impossible goal.

The fact that mathematicians cannot devise a method to factor that large product, despite large amounts of money and research time, is what’s valuable!

But even then your argument breaks down when you consider more complicated cryptographic structures.