LordHumungous

LLMs lack the holy ghost

I developed this for my own personal use. Maybe you will find it useful too!

> Reading everything becomes the default. At a cent per document, a model can read every paper in a field, every record in an archive, every email, or every message in a support queue as a matter of routine

Pretty much already happening

I work at Meta. The security team was recently gutted. 50% were either laid off or moved to data labeling.

What if I tell the model to go commit fraud or crimes and it complies? What if users are having psychotic episodes driven by their interactions with the model?

Just because safety is a hard and messy problem doesn't mean we should just wash our hands of it.

Why not rewrite if code is free

However there are many people out there making the argument that code is free or nearly so. I think the article is directed at them.

Usually all code has an owner though. If I encounter a bug the first thing I often do is look at git blame and see who wrote the code then ask them for help.

Kind of funny ngl

Most people are far more concerned about their livelihood than about abstract notions of beating China in a game of geopolitics. In that scenario the US becomes an isolated economy, and other nations likely follow suit. It will be a poorer, less dynamic world, but most people will choose that outcome over poverty.

That would be an interesting scenario.

If the worst predictions about AI's effect on employment turn out to be correct, then I'd expect to see movements to force government regulation of AI. Particularly if it becomes the case that profits are accruing to a few massive corporations who run the AI.

There is no reason people have to tolerate a technology that is destructive to society, anymore than they have to tolerate companies selling fentanyl at 7/11.

> For apps that run locally—no servers, no cloud costs—subscriptions make no sense anymore.

Did it ever make sense? I always scoffed at the idea of paying a subscription to use a text editor or paint tool.

LLMs finally deliver on the crochety front end dev's dream of writing everything in vanilla JS. Hallelujah.

I have mixed feelings. On the one hand, I totally feel what this author is saying. On the other hand , I love that I am now able to push into areas that I could have never touched before, and complete successful projects in them.

Do people actually have success with agent orchestrators? I find that it quickly overwhelms my ability to keep track of what its doing.

Hi folks,

I created this Claude plugin that allows you to set a target for the percentage of human generate code in your project, and when you are below that target it hard blocks Claude from writing code.

Like many people, I am simultaneously excited about the power that AI gives me, and also fearful of what will happen if I give up coding altogether, both to my competence and to my personal satisfaction. I find that using this plugin allows me to set a balance that falls between pure vibe coding and writing everything by hand.

If this resonates with you, I hope you find some use from this plugin.

Regards,

Matt

How do you pass the bar exam without the ability to reason?

Let me guess, no actual data on DAU, just subjective impressions about the quality of search results?

You put your finger on something that I've been thinking but could not articulate.

Twitter amplifies the voices of the "right" type of person: celebrities, journalists, those verified by blue check.

There's evidence for limited activities, no evidence that foreign action is responsible for the macro trend of partisanship im the US

No that's of girls who have thoughts of suicide, 13.5% say it makes it worse. Ok, is that good or bad? The number is meaningless without a control.

This is the kind of conspiratorial thinking that makes me think we're in a moral panic here.

All I can suggest is look at the actual data and not the reporting on the data.

Also there's the problem of a lack of a control group in this data. If you asked people in an online questionnaire how the Coca-Cola corporation was affecting their mental health what sort of responses would you get?

I don't see any divisiveness nor politics on my feed. It's all baby pictures and vacation photos. Anecdotes are not data.

Nobody seems to care to look at the actual data she provided. It's weak imo. 15% of teens say FB makes their lives worse, 34% say it makes their lives better, and 51% say neither. This is an online poll. Among teens who are depressed, more say that FB/IG has a negative impact. Well ok, a more negative response is exactly what you'd expect from people who are reporting to be depressed.

Then there's the supposed impact of the algorithm on divisiveness. There's almost no data backing this conclusion up, just anecdotal complaints from some European political parties and oh yeah, media outlets like Buzzfeed.

Overall, this data is not compelling and the conclusions drawn from it appear highly sensationalized.

C levels don't sit on the call with engineers. They aren't that dumb. Managers will communicate upward.

My team set up a discord lol