I think lot of Americans overestimate how much China needs them. (Understandable, I suppose, since the UK made a similar mistake vis-a-vis the EU about 10 years back.)
From a Chinese perspective I expect China’s largest customer is China. These days it’s actually kind of wild how many Chinese consumer products aren’t (and won’t be) available at US retailers. And a lot of them are quite good.
For a while now China’s wanted to reduce its dependence on the US for a variety of reasons. And undermining the US tech industry, whose products the US government likes to use as a cudgel, may serve that goal quite nicely.
(Speaking as a lapsed participant in a zen community. Opinions are my own, so others may disagree.)
The challenge for these “Zen and my job” types of work is that much of Zen practice is (deliberately) a bit of an inkblot test. If there’s any real Big Truth to it, it’s that it’s all one big muddle that everyone has to muddle through for themself. Unlike in many Western religious traditions, there is no presumption of univocality - even within the canonical literature, every opinion is understood to be the author’s alone and represents their own personal attempts at muddling things through. It’s just that some people and works are regarded as being particularly worth a read.
But that’s never explicitly stated. Which can be particularly confusing to people who grew up in religious traditions that do presume univocality and a greater degree of spiritual authority. Hence these works by people who see one particular bit that clicks for them and conclude, “Ah, this is what Zen is really about, I must go tell others!” without realizing that perhaps they just stumbled across an inkblot that they found to be particularly evocative.
It is. The next post in the blog expands on the idea in concrete terms. The author doesn’t explicitly say so, but based on the shape of it I would assume all we have right now are the introduction and first full post of what is intended to be a multi-part series.
This is nothing new. For about 30 years now Microsoft has been constantly repeating various flavors of this “make it so a thing can automatically and silently run programs as soon as it touches your computer” thing. It’s always done in the name of user convenience. It always ends up being a fiasco. I don’t know why they keep doing it, it’s not like the exact same PHB keeps making the same decision over and over for 30 years. It’s probably one or a combination of the many well documented flavors of stupid that are deeply baked into the company’s organizational culture.
(And before the inevitable response, no this is not defending Microsoft. Pointing out that an organization’s culture is too deeply, chronically stupid to avoid opening the exact same obvious and gaping security hole over and over and over and over again is not the same as saying, “it’s fine, actually.”)
Yes. Just this evening I had to help my mom. She was convinced a web form she needed to fill out was broken.
The problem was a required input control was so heavily styled that it was genuinely difficult to recognize it as interactive. Possibly impossible for my mom, whose eyes aren’t what they used to be.
And, in place of traditional validation design language, they decided to just make the submit button invisible until the input validated.
The worst part is, someone is probably quite proud of the extra day or two they spent making sure that UI was broken.
I think that’s the response the “it’s never a real problem until suddenly it is” observation was trying to head off.
There’s always someone who doesn’t want to take the advice of all the people who’ve been burned because “well it’s never happened to me.” I used to be that someone myself. Until the day it happened to me.
That is the nominal motive, but I think that the real motive is that it makes it easier for average developers to pile on more complexity.
Back in the day we just accepted technical constraints that didn’t prevent us from doing anything mission critical. So the UIs weren’t paragons of personal expression, probably there was only one database which didn’t have any scale out features at all, and nothing scrolled infinitely. And I do think that tended to make things easier to fit into your head, because the number of distinct APIs and module interactions you needed to understand was much smaller.
But also none of it was integrated with Alexsirtanapilot, so everyday life was an endless misery.
I’ve mainly handled it by pushing my team to be extremely conservative about what dependencies we take, especially if they pull in scads and scads of transitive dependencies.
This elegantly mitigates three problems in one go: update churn, dependency hell, and supply chain attack surface.
It also, frankly, tends to make the code easier to understand. I’m not a huge NIH person but I do have to say that a lot of packages these days tend to encourage ways of doing things that are unnecessarily complex. More than once I’ve replaced a dependency with homegrown code and reduced LOC in the same commit.
I would argue that, over the time scales at which Von Neumann probes would hypothetically spread, “report home” may be a useless or even wasteful requirement. Even if somebody were still around to hear the message, what is the likelihood that they would still be listening? Or be able to interpret it?
Of course not. That isn’t what the article was trying to say.
The thesis was, “I would like to suggest that the hardest part is the one that gets a single sentence: ‘mines local material and builds a copy.’” And so naturally the points in the article are only trying to support that point.
The closest we get to what this thread is talking about is the concluding remark on the Fermi paradox. Which doesn’t rest on the idea that it’s a practical impossibility; just on the suggestion that it may be hard enough that we can’t just assume civilizations that are in principle capable of building them are likely to actually do it.
It possibly goes a little further, though, since we’re talking about multi-step agentic loops here. So in addition to next token prediction there are of iterations of collecting input from outside sources.
I’m still hesitant to interpret this as “thinking ahead” without at least seeing some more back-and-forth in the literature first, though. This just seems like one of those spots where it makes sense to give other researchers some time to come up with additional hypotheses to explain the observations instead of focusing on the first one anyone proposes in isolation.
I don’t really like much about C++ anymore, but I still enjoy reading C++ articles and listening to C++ podcasts, and I would consider it beautiful. Oftentimes the things I dislike about it are also the beautiful things. The term “beautiful mess” seems appropriate.
It’s a bit like a well-kept Victorian home. The amount of work, money, and dealing with discomfort that goes into maintaining one isn’t something I really want to experience for myself. But the amount of skill and craftsmanship that it takes to preserve one is still impressive, and I have to appreciate the respect for history and the care that goes into balancing it with modern concerns.
And talking to people who do live the life is always a great learning experience.
I’m not sure how much it makes sense to pick this nit for a public opinion survey of this nature. The survey is being sent to people who mostly won’t think of the term in such precise ways, and even in social sciences it’s considered poor form to try to measure more precisely than your noise floor permits.
That said, I would assume most respondents have a more popular conception of the term. That’s going to be inherently a little fuzzy, but implies implies sites like X, Facebook and TikTok count, that Reddit is marginal, and that more “oldschool” things like webforums, chat services and even Hacker News are out.
I’ve been working in AI - and specifically NLP - since 2003. I am no stranger to how weird quirks can sneak into overparametrized models, nor am I a stranger to how good humans can be at inferring meaning where there is none in specific language model behaviors. So, yeah, I am inclined to assume non-teleological causes are more parsimonious than inferring the presence of a strange loop, because that continues to be the winning bet. Even for generative LLMs.
The token saving is oversold, from what I can tell so far. These days output tokens are just the tip of the iceberg.
If anything the real value is it saves my brain from going into power saving mode by lunchtime because I haven’t spent the day reading pages of output when a sentence or two would do.
It sounds like you are trying to understand LLM behavior using a mental model that inaccurately personifies the stochastic parrot.
A more parsimonious explanation is that this term got more-or-less randomly boosted by the reinforcement learning loop because there was nothing in the training data to discourage its use.
And like any good corporate buzzword, it’s merely a simulacrum of precise technical jargon. The way Claude uses it is clearly wildly polysemous if not outright ambiguous.