VMG

Geübergegenbeispielt

Isn't this just an effect of what the LLMs are RL'ed for? Solving short-horizon tasks.

I assume one can't benchmaxx multi-year long efforts, clean architecture, taste etc as easily as these "make tests pass" tasks

Here’s my dystopian sci-fi scenario:

As prediction markets already show, forecasts can influence the outcomes they are trying to predict.

What happens when these models become extremely accurate and widely trusted? A forecast like “Will there be a war between countries A and B?” may itself affect whether the war happens.

If the model says there is a 1% chance of war, little changes. But if it says 90%, governments, markets, militaries, and the public may react: capital flees, troops mobilize, diplomatic trust collapses, and each side starts preparing for the other side’s preparation. The prediction helps make itself true.

The same feedback loop could apply to bank runs, market crashes, civil unrest, elections, and corporate failures.

At some point, the most accurate forecaster may become less like an observer and more like an actor with enormous power over the system it predicts.

well obviously N=1

Crank blog, very skeptical

but just as useless

Empty strings are usually an artifact of lazy developers paying a minimal "empty" value for a type (just as 0 for numbers).

A type like NonEmptyString is a weak defense against that, as a lazy dev can just pass a single space character or something similar.

Is it possible to tell slop from non slop if you were not there when the tokens get emitted? Somebody can just lie and pretend that they were not generated

I must admit I don't really understand what the point of the post-install script concern is.

Usually, you run the actual packaged dependency code at some point anyway, and usually with the same permissions as the install process.

So all of these setup scripts (good or bad) can just move their entrypoint from npm to wherever the `import` or `require` happens.

It seems to me that this is a small stumbling block at best, unless the whole ecosystem moves to a deno-like sandboxed environment. Maybe that is the plan?

I don't know, maybe something about backwards compatibility, maybe nobody can agree on how to do it correctly. It hasn't happened for decades, so I'm not going to hold my breath.

most crypto mining has moved to specialists, even where there were deliberate attempts to make it ASIC-resistant

SETI@Home is a very niche use case

and web browsing still happens by connecting to data centers and server farms, not by connecting to another laptop

Unfortunately, real apps and native tech stacks can not only write data to your SSD, they can usually write data to the user directory however they want and they can read it as well!

Browsers are at least somewhat sandboxed

if there end up being useful workflows where you keep stuff running in the background or overnight that's one advantage

That is not how LLMs are typically used though in my experience

Think of it like having a graphics card at home versus using a cloud gaming stream?

Latency seems to be much more important in that use case

Convince me

1. in order to run LLMs, especially the best ones, you need complicated devices which are expensive

2. if you buy one for your personal use, you are probably not going to utilize it all the time and it will be idle a lot

It seems to me that it will always be more economical that the LLM-running devices are in a datacenter where it is easier to make sure they are always utilized

The problem is that often the program runs into some edge case that requires interpretation, at which point one is tempted to let the LLM deal with the edge case, at which point one is tempted to let the LLM deal with the whole loop and let it do the tool calls

... unless you actually want to edit a change!

I've had mixed results.

Most models don't have a 100% correct CLI usage and either hallucinate or use some deprecated patterns.

However `jj undo` and the jj architecture generally make it difficult for agents to screw something up in a way that cannot be recovered.

blast from the past - peak of UX!

https://de.wikipedia.org/wiki/Norton_Partition_Magic#/media/...

503

Skeptics Guide to the Universe

developers with good taste like Andreas Kling will be able to design entire OSes with coding agents

Not at all if you consider the internet pre-LLM. That is the standard expectation when you load a website.

The slow word-by-word typing was what we started to get used to with LLMs.

If these techniques get widespread, we may grow accustomed to the "old" speed again where content loads ~instantly.

Imagine a content forest like Wikipedia instantly generated like a Minecraft word...

Base64-encoded secret in URL Prevented Detected (entropy scan) Logged

Ok so how does this "Entropy scan" work?

Apparently by defining "bits per character"

https://github.com/luckyPipewrench/pipelock/blob/3021f023b0e...

So I guess converting the secret to pure binary will evade the "entropy scanner"?

Because if it’s worth your time to lie, it’s worth my time to correct it.

https://www.astralcodexten.com/p/if-its-worth-your-time-to-l...

and I expect within the next ~2 years AI tools will produce a better compiler than gcc

and the "anti" crowd will point to some exotic architecture where it is worse

Step 2: outlets slap this disclaimer on all content, regardless of AI usage, making it useless

Step 3: regulator prohibits putting label on content that is not AI generated

Step 4: outlets make sure to use AI for all content

Let's call it the "Sesame effect"

have you read the linked page?

However, since immunizations are given to about 90 percent of children less than 1 year of age, and about 1,600 cases of SIDS occur every year, it would be expected, statistically, that every year about 50 cases of SIDS will occur within 24 hours of receipt of a vaccine. However, because the incidence of SIDS is the same in children who do or do not receive vaccines, we know that SIDS is not caused by vaccines.

the Kaufland ones where I live still have weight sensors which for me completely eliminates the appeal

Telnet is "sandboxed" in that it can only output characters to your tty, however that in itself is quite a powerful primitive.

The ANSI control characters wield power of a huge stack of not very robust code

https://nvd.nist.gov/vuln/detail/CVE-2024-56803

Do you believe that it is impossible to advertise, spread fake news or propaganda via text?

Do you know what the letters in LLM mean?