jesse_dot_id

There are humans behind all of these actions. We're just not holding them responsible for some reason re: hacking into tech companies.

I truly don't understand why people think LLMs eliminate the need for mastery, or that they limit thinking. I have a lot more time to think when I use a good harness. It handles the pointless toil and I handle the thinking. I'm just thinking about different things. Instead of trying to coax the correct answers from a search engine, the LLM is doing that for me and parsing the correct output, so I'm not spending hours upon hours wading through detritus. While it's doing that, I'm using that time to think more about the problems I'm solving, the UX, etc.

I'm not sure what to tell you if you think the LLM solves for novel thinking.

Imagine saying this about a graphing calculator, because that is essentially what the LLM is for knowledge work. You can either be dumb and spell 8008135 or you can get an education and make it sing. They definitely want us dependent on their products for life, because that's what capitalism is, but dumb kids won't be the profitable ones.

It's because it's incredibly stupid.

There's been a Chrome CVE like every week ever since it came out.

Did they give themselves the existing guardrails?

Programming?

Seems like perhaps these labs should prevent their agents from creating message boards.

Instinct. I think it drives UX. When I am designing an interface, I'm not just parroting something I've seen before. I am calling upon the breadth of my human existence, my professional experience, my empathy, to think about how another human will be using the applications that I create. It paints everything I touch.

LLM will never be able to do that. It can clone things that humans have created in the past but it can't create novel solutions that a human will enjoy without a human at the wheel iterating through prompts until it's great. AGI doesn't mean 'good enough' to me. It means better.

No, they're pretty clearly gaming benchmarks.

They're gaming benchmarks HTH

AGI to me means intuition and I don't think that's ever going to happen with a LLM.

Press X to doubt.

I cite it because we're barely scratching the surface of our own world, which is one of a potentially infinite number of worlds that is encapsulated by the word "everything."

Why would we map the seafloor is an insane question that immediately invalidates all other opinions that you may have, unfortunately. The human embodiment of that meme with the pickaxe guy walking away from a diamond strike if only he had swung once more.

It's so hubristic to assume that our generation is the one that will discover the answers to everything.

Probably never. We've only mapped like 29% of the seafloor, actually explored like %5 of it, and we've barely pierced the earth's crust.

Staying silent is unfortunately how fascism festers.

I do not trust any of them. I do not trust xAI the most.

Same experience on 5x.

Not for enterprise. Can't trust the company behind it with my data.

I love em dashes because they are kind of a wildcard. When I read this same sentence, I interpret this emdash as an ellipses and not a comma.

I had just assumed this model would read differently due to watermarking.

Impressive article. I read most of it without skimming and that is high praise from this AuDHD nerd.

It's definitely suspicious. My money is on inside job — perhaps someone not fond of the current administration's handling of things. In my experience, these types of vulnerabilities are well-known and kept in the back pocket of engineers that have a lot of experience maintaining the equipment. Everywhere I've worked has been deeply vulnerable in some way, all of the engineers knew it, and often joked about it or were counting the days to getting pwned because companies never like to spend money to address the issue. I'm sure the military is no different. Probably worse.

Still a no from me until these things can display any degree of intuition, and even then it's probably still a no.

You're asking for trouble if you hook an autonomous agent up to anything that you care about. You're insanely naive if you give it access to everything.

OpenAI can't even monitor its own shit properly with teams of well paid engineers, and you are one novel prompt injection technique away from your entire digital life going up in smoke.

Also...

>Today we released by far the largest update in the history of OpenClaw. It was built by 933 contributors, including 569 first-time contributors, and is composed of over 16,000 pull requests.

This is not the flex you think it is. That's horrifying lol

Have you ever seen really good technical writing versus really shitty technical writing? Which do you prefer? Why? The LLM is reading instructions. The quality and design of the instructions you give to it matters deeply. If you don't think so, I don't particularly care, but you can't really approach a LLM like an easy button.

It's like a graphing calculator. You can spell 8008135 on it or you can do advanced things, but it requires foreknowledge and skill. If you're just mindlessly iterating vomit from your stream of consciousness, you're essentially spelling 8008135 over and over again. That feels novel now. It's not going to remain novel and you will stagnate as things progress.

Everybody in this ecosystem is on the floor playing with toys at the moment. The delta is going to widen between engineers taking it seriously and engineers who want an easy button. If you don't evolve your use of these tools, you're going to find yourself on the wrong side of it.

Yes? If I've developed a prompting style over the past four years that produces efficient and secure code bases with novel features, I consistently out-perform my colleagues using the same tools, and it is how I make money, then I don't see how it is not a trade secret.

I consider my prompts and prompting style to be trade secrets. I've developed a very efficient and productive workflow for myself over the years without tutorials, without books, etc.

Anthropic is making very strange and very opinionated decisions far too often for my liking. It's really making me second guess my continued use of Claude Code.

Zoinks I mangled that. Late edit:

I've known this for awhile, since at least 2020. Specifically when they told a documentarian he could film and then threatened litigation if he released his film — after he had already shot it. He had to painstakingly blur literally all Burning Man footage except for the guy he's following around.

I've known this for awhile, since at least 2020. Specifically when they told a documentarian he could film and then threatened litigation he release his film after he had already shot it. He had to blur literally everything at Burning Man but the guy he's following around.

https://www.blurringman.org/blurring-man

My company has a great culture and also AI.

A lot of crypto bro grifter crossover in AI.