HackerTrans
TopNewTrendsCommentsPastAskShowJobs

lelanthran

13,470 karmajoined 7 jaar geleden
https://www.rundata.co.za https://www.lelanthran.com

Submissions

Don't use localhost:3000, use your own custom domain

idiallo.com
4 points·by lelanthran·11 dagen geleden·1 comments

Story Points Revisited

ronjeffries.com
5 points·by lelanthran·vorige maand·0 comments

LLMs Are Not a Higher Level of Abstraction

lelanthran.com
166 points·by lelanthran·2 maanden geleden·156 comments

LLMs are not a higher level of abstraction

lelanthran.com
4 points·by lelanthran·2 maanden geleden·10 comments

AI's Mainframe Moment

mjeggleton.com
3 points·by lelanthran·3 maanden geleden·1 comments

I learned Unity the wrong way

darkounity.com
181 points·by lelanthran·3 maanden geleden·115 comments

C and Undefined Behaviour

lelanthran.com
18 points·by lelanthran·5 maanden geleden·12 comments

More States Are Taking Aim at a Controversial Early Reading Method

edweek.org
3 points·by lelanthran·5 maanden geleden·0 comments

Security vulnerability found in Rust Linux kernel code

git.kernel.org
37 points·by lelanthran·7 maanden geleden·19 comments

Where's the Shovelware? Why AI Coding Claims Don't Add Up

learnui.design
3 points·by lelanthran·9 maanden geleden·0 comments

Why SaaS Pricing Pages Fail (and How to Fix Yours)

dnsk.work
1 points·by lelanthran·10 maanden geleden·0 comments

comments

lelanthran
·1 uur geleden·discuss
> The number of compiler runs doesn't matter as much as the total elapsed time it takes to finish the task. In just about every test we ran, LLMs are faster at building in Python than Haskell.

How much faster? IOW, what's the difference (in minutes/seconds, not in percentages) between vibing Haskell and vibing Python?
lelanthran
·19 uur geleden·discuss
> Did Emacs really give YOU any capability you didn't have before?

`M-x docter` is something I never had before.
lelanthran
·19 uur geleden·discuss
> Honestly, at this point isn't not exposing our kids to AI just setting them up to fail in the future?

The whole point of AI is to enable unskilled people to perform skilled tasks.

In your mind, what skill does a person actually learn from getting onboard AI usage early? It's the other way around; those who come to it late (when the models are more capable) are going to have a shorter learning curve.
lelanthran
·gisteren·discuss
Ideation.

We no longer meet, we jump on a call.
lelanthran
·gisteren·discuss
> overuse certain tropes and patterns because those patterns are commonly represented in human writing.

That's factually untrue. Give me a link to a pre-2020 piece of writing that sounds like an LLM.
lelanthran
·eergisteren·discuss
> . You should shift your technical insight to a higher abstraction level, where the AI cannot help anymore.

As a programmer, you also had to work at that higher abstraction level anyway.

It's a myth that you are "moving" up a level; you were always at that level, just not exclusively.
lelanthran
·eergisteren·discuss
> do you mean my enjoyment from building things? I'm genuinely confused by this response.

I'm not surprised - your GGP comment indicated that you are more interested in the destination than the journey (you enjoy the output more than the process of crafting that output).

Nothing wrong with that - lots of programmers are interested in the final deliverable and don't really care about how the sausage is made, but you're reading a comment from someone who makes the sausage.
lelanthran
·3 dagen geleden·discuss
> Honest question: couldn't this be solved by setting the authorization level of the agent the same as the user that prompted the question?

No.

> In this post's example, the agent would be limited by the author's scope inside the organization and, therefore, be incapable of exposing any unwanted file.

That still allows prompt injection to exfiltrate the authors files. That's the whole exploit - files that the author has, that he doesn't want exfiltrated.
lelanthran
·3 dagen geleden·discuss
> Partially, you could still deploy the AI in an isolated envirnoment. If there's nothing to access, there's no prompt injection.

If there's nothing to access, there's only limited value in using an LLM in the first place.

If your LLM is prevented from accessing anything other than the prompt, the only use is interactive use by the user; no automatic work done on any workflow items.
lelanthran
·3 dagen geleden·discuss
> You can just make the tool calls restricted/scoped to whatever the calling account has access to (or in this case the repo)

This is a fix for the harness, not the model.

As an analogy to SQL, this is like "fixing" SQL injections by having JS on the frontend escape/sanitise the values sent to the backend, while the backend does not use parameterised statements.

The harness is the front-end, the model is the backend. There is no way to currently fix the backend with parameterised prompts.
lelanthran
·3 dagen geleden·discuss
> But then some genius discovered that if you ran a wave form through a clipper which gave a sequence of 1s and 0s, running that produced quite a credible voice sound.

I'd like to know more about this, actually.
lelanthran
·4 dagen geleden·discuss
> Oh yeah I forgot 5 BILLION dollar in revenue is a signal that the market does not want your product

Doesn't matter how big the revenue is if the market is not interested at the price you are selling.

I mean, by your logic, if I sell a dollar for 64c, and do $5b revenue, that's an indication that the market does indeed want the product, but not an indication that the market wants the product at the price you need to sell at to stay in business.
lelanthran
·4 dagen geleden·discuss
> Oy vey maybe this kind of money grubbery is not good for our society.

No, it's a signal from the market that the product being sold is not wanted by the market.

They've mismanaged all their IP pursuing that yummy subscription revenue. Turns out gamers really don't like to buy subscriptions. As a poster downthread pointed out, the games that are not always-on and subscription-based are doing fine. It's the recent AAA model of subscription that is bleeding money.
lelanthran
·4 dagen geleden·discuss
> So first of all, they are big! Secondly they are not at a loss. They just have a "thin, non-growing margin".

At a roughly 3% profit, they are barely breaking even and have no money to invest in staying current. It's not a sustainable business.
lelanthran
·4 dagen geleden·discuss
I think the future will have to include specialised host boards for memory chips.

What I actually want is an FPGA board with a very large number of DDR3/DDR4 RAM slots arranged in banks (2, 4, 8 or even more banks). I want an FPGA board that can hold 1TB of DDR3/DDR4 RAM.

The throttling point right now is not RAM, it's bus speed. Having different busses for banks of RAM alleviates that.
lelanthran
·4 dagen geleden·discuss
> I understand the arguments for a margin collapse, but I don't see any historical analogues.

How is this the top comment? It lists all the outliers and ignores thousands of instances where fat margins caused a collapse.

I mean, just what Linux did to the dozen or so fat-margin unix server companies is already a longer list of collapsed companies than provided in this comment.
lelanthran
·4 dagen geleden·discuss
Sure:

1. A primary meaning, "Further" means "revenue" (or "profit"). You and your AI together can sit and create a competitor to Windows 11, do the marketing copy, a sales strategy, feature testing etc, but without a team that product isn't going anywhere.

2. A secondary meaning, "Further" means "Stamina". You + AI can create everything above, but without a community (whether paying customers or free users), the product dies when your interest in it dies, even if it is still making money!
lelanthran
·4 dagen geleden·discuss
Ah yes; thanks. the xcancel one was what I was thinking off.
lelanthran
·4 dagen geleden·discuss
Right, thanks.

I looked at the format. I think you're mostly on the right track, but I also think that a better candidate might be to simply use (and augment, where necessary, such as for styles) the org mode format: It can do all the stuff you have, but also things like checkboxes, calendars, and more.

As a bonus, both people and agents already know the format so there is no need to have a skills file. For example, the following prompt on Gemini WebChat (hardly a good model):

    Give me an org mode file to show a PERT (Project evaluation and Review Technique) diagram, with a calendar below the diagram allowing me to see the current year. Create a hierarchy of tasks that have to be done using  checkboxes and collapsible sections to mark tasks/subtasks as done. Below that, give me a table of all the terminal tasks that need to be completed with task/subtask name, starting date, estimated ending date and the resource assigned to it.

    Finally, at the end, produce a gantt chart as a mermaid diagram for the sample project.
Produced a working file with tables[1], diagrams, calendar, checkboxes in a single file that Emacs rendered properly. Org mode can export to every format I ever needed (LaTeX, html, pdf). I once even had the resulting HTML conversion contain animations written in Javascript :-)

Maybe all you need to code for agents to write is a web-based viewer for Org Mode syntax?

Look at it this way: right now if I wanted what smalldocs does (i.e. ask the agent to generate any of your examples), I can ask the agent "do $FOO, generate org mode", and without a single additional skill/claude.md/agents.md file, get exactly the result you got from smalldocs.

I think maybe testdrive Emacs daily for a month; it would open your mind to the possibilities available[2]. If anything more is needed (like I wanted to put in JS in the HTML output), it can do it. If Emacs cannot do it, my agent can write an EmacsLisp function that will do it.

At the end of the day, when even a poor LLM can do what smalldocs does but without any additional .md files or context, I think maybe your solution might be over-engineered.

----------

[1] Org mode tables work exactly like spreadsheets, in that they can contain formulas.

[2] Think of it this way - when I needed multimodal documents, because I already knew Emacs, I just used that. When you needed multimodal documents, you vibed a whole new product into existence.
lelanthran
·5 dagen geleden·discuss
> Perhaps. I think AI changes the equation here. Honestly, AI changes what "solo developer" even means.

I disagree; it's even more obvious with AI that, with AI, a solo dev can go even faster, but still, with AI, you need a team to go further.