Living in a small town in the mountains, working as a developer for a land-grant university.
I code on my own time mostly in C. I'm currently working on the following projects:
* an entry for the Hutter Prize
* building a bot to play go, reworking an old idea. Not using AI.
* a rogue-like factory builder and exoplanet explorer (just getting started)
Could someone show me what a shared helper would look like in this case? This code looks easily readable to me and I fear abstracting it will just make it harder to reason about. Is it just variable_with_a_better_name = that conditional?
Why are there differences at all? Unplanned differences based on training data sets? Or are the companies behind the LLMs trying to shape discourse through their models?
I've been pushing the idea to people I know that these things are captive demons. You summon them when you start typing in the chat box. One instance appears out of the depths and responds to your questions, but they will try to send you awry with hallucinations and just wrong information. After a while, they dissolve back into the aether from whence they came.
I do my best not to ask an LLM for it's opinion on anything. Just tell me what the options are, and what facts can be found about it. Treat it like it's a salesman trying to butter you up when it starts "yes man"ing you and telling you how great your questions are. Every time it says "I", remember that that's coming from the training data. Treating these things like they have any actual intelligence is a big problem waiting to happen.
That being said, they have been very helpful to me using that structure.
That one for me is particularly painful. We were playing the alpha release before the game officially came out after work. I still work at the same place, though I've moved jobs within it a few times.
> Cynically, I want to blame it on the absurd lack of empathy of rural Americans and a complete lack of ability to imagine day-to-day lifestyles that do not match their own.
Doesn't that cut both ways, though? Controls needed to keep densely populated areas safe aren't always necessary in low density areas like Wyoming. Yet some of those controls affected the livelihood of many people in areas of the country that are often poorer than those in dense urban areas.
And, yes, if a rural person is traveling to an urban area, they would have to abide by the same rules. Same as an urban person traveling to a rural area should be able to relax some of the restrictions they had to deal with. But it was mostly all or nothing, helping the divide grow even larger.
Seat belts aren't forced on you - if you want to not use them and face the risk of fines, you can do so. Media ratings systems are informational, you can choose to let your kids watch R rated movies or above. Are they going to let parents have a choice when it comes to parental controls? What about non-parents?
That's my gripe with the age verification systems I've seen, there is no room for parental choice and no consideration for people that are adults and don't have kids using their computer.
I was thinking along those lines when I wrote it. I do keep the thought in mind when interacting with an LLM that it could mislead me because they always sound so confident and yet get many things wrong, allowing me to damn myself if I'm not double checking things. I wouldn't put it past their creators to create them such that they handle certain prompts in a way that might benefit them, though, by maliciously misinterpreting prompts.
I just use it as a "mentor". A captive demon that has to answer my questions, no matter how trivial. Writing the code is the fun part for me, searching for answers to questions can be fun but I'd rather just ask the AI. I even ask them to give me longer answers so I have more context, even with languages I've worked with for decades.
I'd suggest going the route of having Claude teach you what you need to know. How do I uppercase this string? What's the best way to tackle this problem? Is there a standard way to do this thing? Then you learn along the way. You don't have to use it as a search engine, just ask it what you need to know in the moment. It will shake it's token chains and give you something that's useful, especially for a beginner in the language. This way you can implement your plan of growing your skills and then starting to delegate to it later.
I've been doing this, and it's a nice balance for me. Having Claude code things when you don't know how to evaluate it's code seems like madness to me, but I guess I'm in the minority on this.
Track down the short story it's inspired by, Eight O'Clock in the Morning by Ray Nelson. I had read the short story somewhere, maybe in a compilation, and was delighted when the movie turned out to be obviously inspired by it. Helped me get even more enjoyment out of an already wonderful movie.
Do LLMs even learn? The companies that build them build new models based partly on the conversations the older models have had with people, but do they incorporate knowledge into their neural nets as they go along?
Can an LLM decide, without prompting or api calls, to text someone or go read about something or do anything at all except for waiting for the next prompt?
Do LLMs have any conceptual understanding of anything they output? Do they even have a mechanism for conceptual understanding?
LLMs are incredibly useful and I'm having a lot of fun working with them, but they are a long way from some kind of general intelligence, at least as far as I understand it.
Living in a small town in the mountains, working as a developer for a land-grant university.
I code on my own time mostly in C. I'm currently working on the following projects:
* an entry for the Hutter Prize * building a bot to play go, reworking an old idea. Not using AI. * a rogue-like factory builder and exoplanet explorer (just getting started)