staticshock

The pelican on a bicycle test is specifically about generating an SVG, fyi, not a raster.

I suspect this article got upvoted partly for the headline & content, and partly because of who said it. Lots of people already have a good feel for the AI/journalism niche Simon Willison occupies, and can use that background knowledge to read into his words more productively.

I think it lands differently for different kinds of prose (e.g. informational, persuasive, etc.)

Persuasive prose, in particular, which is probably the majority of the things that get posted here, is less persuasive to the extent to which it includes obvious "AI tells". Even in the cases where AI is more articulate, (1) the emotional weight of the text feels manipulative when it is clear that the emotions were, at best, "vetted" by a human as opposed to having been produced by one, and (2) I think we have a built-in "bullshit-proximity" sensor, and have recently been trained to expect AI to be more willing to engage in bullshit than humans due to their obsequious disposition. (For the record, I think humans are also full of bullshit, so this one's more of a toss-up.)

Low-tech users don't give a damn if something has the guts of an "app" or not, they care about having a thing on the home screen they can click.

Businesses have the incentive to give their users that low friction experience (at the point of need) using already familiar rails (i.e. "install app from app store").

The makers of both iOS and Android treat the ability to "bookmark" a web URL onto your home screen as a power user feature that requires navigating through complex, technical-sounding menus. Does it have to be like that? Of course not. They just have a business interest in pushing users away from the open web and towards their walled gardens.

--

Mind you, I'm not saying, "advertising doesn't play a role in this". A clump of well aligned motivations is obviously going to be more powerful than a single isolated motivation. But let's not forget that apps built for non-technical users, which—I cannot stress this enough—IS MOST USERS, benefit greatly from lowest common denominator solutions where they never feel like they have to learn anything to get going.

We only come to understand idea quality via evolutionary fitness tests, which means we have to MAKE things in order to find out if they're any good. Everything you make is gonna do hill-climbing in a [potentially unstable] environment in search of an ecological niche. This is true of species a la Darwin, but it's also true of products & ideas.

Anything we make gets to participate in this evolutionary process, even if it sounded dumb at the outset. In fact, doing dumb things is how we get "unstuck" from outdated ways of thinking.

The way I understand doom scrolling (and all other forms of passive engagement, e.g. youtube, reddit, hacker news, etc.) is that when we dip below "baseline" mood (be it through stress or any kind of adversity), quick hits of stimuli momentarily bring us back up to that baseline.

At the same time, access to abundant entertainment establishes in us an unrealistically high baseline that we're destined to fall below whenever we're not actively "plugged in". When you fall below baseline, unrealistic though it may be, you feel lacking, like an addict, and you respond to that feeling exactly like an addict.

It's all too easy nowadays to forget that boredom is good. Adversity is good. Having the time to sit with your thoughts is good. We grow stronger through any form of perseverance, and weaker through any form of surrender.

Think of the entirety of the context (the full thread of conversation, all tool call output, etc.) as one message that's been submitted to the LLM all at once just so it can generate the next token (which is just the next word or even syllable.) Once that token is generated it is appended to the context, and the entire context is once again used to generate the next token. Keep doing that until the entire response is generated. The larger the context, the more stuff the model has to pay attention to as it generates each next token.

To use an analogy: imagine your friend is the author of an unfinished book. They die with 19 chapters written, and on their death bed ask you to write the 20th chapter. Assuming you're up to the task, you can only do this well if you take the time to absorb the entirety of what's been written so far.

This is how LLMs work. Context caching is an optimization on top of this, but it has its limits.

At scale, specs can only be vital to the degree to which their conformance testing is automated. Good specs should use a formal, runnable verification language. Otherwise you'll accumulate specs that are right when they ship, wrong in subtle ways 3 months in, and wrong in glaring ways 6 months in. AI doesn't change this dynamic, it amplifies it.

There's a subgenre of dystopian sci-fi where the premise is that reality in general is destined to be eclipsed by matrix-like hyper-realities, and that people will vastly prefer those and cede reality to whoever's left.

I guess the way this could work itself out is that if you prefer a hyper-reality, your genes do not pass on, and someone else's do, and within some number of generations we bounce back in response to evolutionary pressure.

I learned a fun fact in a recent interview of David Reich (by Dwarkesh):

Every mutation that can occur does occur. There are eight billion people in the world. There are maybe 30 new mutations every generation, so that’s 240 billion new point mutations every generation. There are only three billion DNA bases in the genome, so every mutation that can occur does occur about 100 times every generation. We’re not mutation-limited anymore.

https://www.dwarkesh.com/p/david-reich-2

We can talk to each other like adults, the key is understanding what the goal of any given conversation is. Truth-seeking is just one possible goal among many.

Part of becoming an adult is learning how little most people care about that particular goal, and how big the buffet of alternatives is: creating shared meaning, understanding each other's values, building trust, giving/receiving emotional support, processing grief, etc. (Think of this as an upfront taxonomic exercise, followed by lots of in situ calibration exercises.)

Even for something like "decision making", which, naively, one might assume should be grounded in facts, a lot of the "facts" wind up being fuzzy and subjective. This is baked into the social fabric.

Agreed. Change makes people uncomfortable. The nature of the change doesn't matter; the transition itself is the root of the discomfort.

When things are stagnant, we gradually optimize our lives towards a low energy state and overfit to our exact circumstances. When a change in circumstances reveals past optimizations to be wasted work, it kick-starts the four stages of grief over the loss of that low energy state.

Don't let your guard down. Tricking Opus 4.6 is not impossible, it's just still an active research frontier. Once the right incantation for any specific model is known, it'll be weaponized.

There was an excellent article on the front page recently about role confusion, which highlights just how just far models have to go on this: https://role-confusion.github.io/

Except that is not what's happening. The article clarifies something that is misleading if you interpret the headline in isolation: "high-end M6" means "the high-end variants of the M6 line", not "the entire M6 line".

Arguably, npm does one thing, but it does it poorly.

My favorite of these has always been the USA PATRIOT Act, or the "Uniting and Strengthening America by Providing Appropriate Tools Required to Intercept and Obstruct Terrorism Act". And that one predates LLMs, so… sheesh.

While I'm sold on the fact that modern nuclear can be built & administered safely in the face of natural disasters, and is a net good environmentally, I'm worried about corporate cronyism's corrosive effects on safety (a la Fukushima) and future instability in the form of cornered animals (e.g. Putin, Trump) acting erratically by bombing civilian infrastructure.

For the uninitiated: https://en.wikipedia.org/wiki/Torment_Nexus

Everything is search.

Software development is search through the space of useful/interesting automations. Business is search for product market fit (at the intersection of expertise, capital, problem, etc.) Writing is search for lossless, efficient idea transfer.

AI software development is more search. If we search more, will we find a bunch of garbage? Hell yes. We'll find a TON of garbage. That's not new, though. The world has been writing way more books than you'll ever read, recording more music you'll ever hear, filming more television shows than you'll ever get to watch, etc. A lot of it is garbage, but the good stuff stands the test of time and rises to the top, and I'd rather live in a thriving, flourishing world full of all these things, because there's more cream-of-the-crop when there's more everything.

It's evolutionary fitness operating in the space of ideas. I agree that "maybe later" was indeed a useful mechanism, and maybe even a local optimum in the development methodology search space (which recently experienced a major earthquake!), but evolutionary pressure will bring it back into existence in some form sooner or later.

An agent you can sue would count as a legitimate third party entity. Suing today's agents won't get you very far.

Feels like an unfair judgment of someone else's values, which have no obligation to match yours.

Are you saying that a better way to measure perceived time is something like "1 year, 2 more years, 4 more years, 8 more years," starting from birth, and maybe call each of those increments a "log year"? I like it.

I guess the "natural" base to use to get the "right" number of increments is a pointless exercise, since it ultimately bottoms out in the question of "why is a regular year as long as it is?", but if we assume a base of 2, I'm currently in my 6th log year, and hope to die comfortably into my 7th. Actuarial odds are >80% in my favor.

Love it!

Bug report, if you're the author: I can't delete periods.

Feature request: I wanna be able to move periods from one layer to another.

"can be" ≠ "must be"

A pattern I like for CLIs is that by default each command runs in dry-run mode, and only with `--commit` is it allowed to do dangerous things. Kind of like `git clean` vs `git clean --force`, except that `--force` feels like bad names for the distinction. Likewise, `--dry-run` implies that the command does the dangerous thing by default, which is bad. `--commit` gets the balance right, it sounds right, and it's sufficiently self-explanatory.

(Oh, and there's no shorthand, like `-c`. It's `--commit` or bust.)

That song is etched into my memory and bring back a flood of great feelings. Daily Motion is giving me a playback error, but here's a copy from youtube: https://www.youtube.com/watch?v=xb_PyKuI7II

Contrarian thinking can be great because it taps into the intuition that the masses are mostly followers who can be led anywhere, not critical thinkers who've deeply examined what they believe. Being contrarian, then, is akin to staking out a new leadership position.

The space of contrarian ideas is vast, and most of them are probably bad, but, nevertheless, the willingness to hold unconventional, internally consistent views should be celebrated, because it increases diversity of thought. Our collective hive mind grows stronger through heresy.

However, I like my heresy with a splash of axiomatic precision, which is sadly lacking in this article.

The value of such a benchmark, to me, would be, "what is peak performance", not just "what is mid-tier performance". Also, possibly, "what's the per-dollar performance". Time and money permitting, I'd really want to see your benchmark extended to the large reasoning models.

AGI is 100% possible, even if the current breed of transformer-based models are not it, and even if silicon is not it. There's nothing special about human brains that we won't eventually be able to match (and then exceed) in vitro. We are living proof that intelligence can be built out of matter, and that human-scale intelligence can run on 20 watts. It's not a matter of if, but when.

Feels to me like idealism crossing into realism. OpenAI could be the next Google, or the next Facebook, or the next… I don't know, Netflix?

All those companies (and many other large tech companies) have discovered the same arbitrage that older media companies discovered decades ago, which is that we, on the average, are much more willing to pay with attention than with money, even where money would have been the better choice.

Advertising continues to be one of the most powerful business models ever invented, and I don't think that's changing any time soon.

I just looked up "may you live in interesting times" and learned that it is not, in fact, originally Chinese. Per wikipedia:

The nearest related Chinese expression translates as "Better to be a dog in times of tranquility than a human in times of chaos."

https://en.wikipedia.org/wiki/May_you_live_in_interesting_ti...