That's exactly what we used in on usenet (except,without rendering unless you were using a nice GUI reader, not just tin/rtin)
The problem is that that's too many characters to reserve (they all have to be escaped when you want the actual character) making the resulting text look awful in plain text mode.
> What it can't do (as far as I understand): complex layouts, precise typography, embedded binary content, anything that needs pixel-perfect rendering. Am I missing anything? What are the other limitations I should know about?
Multi-level lists, annoyingly, get rendered as code at the deeper levels because of the 4+ spaces from the beginning of the line.
This is a serious and major drawback of markdown, making it good for developers only. The average person does not want to render code.
Remove that one drawback, and it'll get even better adoption.
> But when someone doing something interesting or non-hierarchical talks about their fuckups, we talk about how misguided their intentions are? Seems a little ~~off~~ unfair to me
In that might tome of an essay, where did he tal about how he fucked up? I read the whole thing and it is clear to me that he doesn't think he fucked up.
> He has opinions on how the company should be managed, what the product should be, and how to interact with the community... but he abdicated all the responsibility, didn't provide leadership, and is now complaining it didn't turn out how he wanted it. This is personality problem and business books won't help.
This isn't a personality problem; it's far far worse (or better, depending on your point of view) than that:
FTFA
> We hired two new staff to work on it, and did our best to reconcile what little guidance we got from Leadership with an internal process focused on discussion and consent. By late 2024, the app wasn’t what anyone wanted it to be,
Yup, no surprises there.
> Toward the end of our time at CAS we experimented with sociocracy as a way to organize without hierarchy and coercion,
> how much to disclose in our negotiations with CAS (IMO everything) or whether board members should be required to donate money (IMO no, plutocracy is bad at all times and at all levels)
> I tried to do what seemed like the only thing I could do in a hierarchy
> Accepting a grant without any consultation with staff about how its obligations might be met
...
> Since the exodus, Leadership has improved on some fronts [...] They hired three new engineers for the mobile team that seem both experienced and enthusiastic.
Well, a large grant will let you do that :-/
It seems to me that the problem was not one of personality, but of ideology.
Personality is deeply embedded in humans, ideology is merely adopted.
Author's ideology differed from that of leadership. His personality is probably irrelevant.
> So "200 lines" captures the concept but not the production reality of what is involved.
How many lines would you estimate it takes to capture that production reality of something like CC? I ask because I got downvoted for asking that question on a different story[1].
I asked because in that thread someone quoted the CC dev(s) as saying:
>> In the last thirty days, I landed 259 PRs -- 497 commits, 40k lines added, 38k lines removed.
My feeling is that a tool like this, while it won't be 200 lines, can't really be 40k lines either.
Google is a special case - ever since LLMs came out I've been pointing out that Google owns the entire vertical.
OpenAI, Anthropic, etc are in a race to the bottom, but because they don't own the vertical they are beholden to Nvidia (for chips), they obviously have less training data, they need constant influsx of cash just to stay in that race to the bottom, etc.
Google owns the entire stack - they don't need nvidia, they already have the data, they own the very important user-info via tracking, they have millions, if not billions, of emails on which to train, etc.
Google needs no one, not even VCs. Their costs must be a fraction of the costs of pure-LLM companies.
> OpenAI, Anthropic, etc are in a race to the bottom
There's a bit of nuance hiding in the "etc". Openai and anthropic are still in a race for the top results. Minimax and GLM are in the race to the bottom while chasing good results - M2.1 is 10x cheaper than Sonnet for example, but practically fairly close in capabilities.
> There's a bit of nuance hiding in the "etc". Openai and anthropic are still in a race for the top results.
That's not what is usually meant by "race to the bottom", is it?
To clarify, in this context I mean that they are all in a race to be the lowest margin provider.
They re at the bottom of the value chain - they sell tokens.
It's like being an electricity provider: if you buy $100 or electricity and produce 100 widgets, which you sell for $1k each, that margin isn't captured by the provider.
That's what being at the bottom of the value chain means.
I get what it means, but it doesn't look to me like they're trying that yet. They don't even care that people buy multiple highest level plans to rotate them every week, because they don't provide a high enough tier for the existing customers. I don't see any price war happening. We don't know what their real margins are, but I don't see the race there. What signs do you see that Anthropic and Openai are in the race to the bottom?
> I don't see any price war happening. What signs do you see that Anthropic and Openai are in the race to the bottom?
There doesn't need to be signs of a race (or a price-war),only signs of commodification; all you need is a lack of differentiation between providers for something to turn into a commodity.
When you're buying a commodity, there's no big difference between getting your commodity delivered by $PROVIDER_1 and getting your commodity delivered by $PROVIDER_2.
The models are all converging quality-wise. Right now the number of people who swear by OpenAI models are about the same as the number of people who swear by Anthropic models, which are about the same as the number of people who swear by Google's models, etc.
When you're selling a commodity, the only differentiation is in the customer experience.
Right now, sure, there's no price war, but right now almost everyone who is interested are playing with multiple models anyway. IOW, the target consumers are already treating LLMs as a commodity.
Email seems like not only a pretty terrible training data set, since most of it is marketing spam with dubious value, but also an invasion of privacy, since information could possibly leak about individuals via the model.
> Email seems like not only a pretty terrible training data set, since most of it is marketing spam with dubious value
Google probably even has an advantage there: filter out everything except messages sent from valid gmail account to valid gmail account. If you do that you drop most of the spam and marketing, and have mostly human-to-human interactions. Then they have their spam filters.
I'd upgrade that "probably" leak to "will absolutely" leak, albeit with some loss of fidelity.
Imagine industrial espionage where someone is asking the model to roleplay a fictional email exchange between named corporate figures in a particular company.
> To be more specific, Sound Reproduction Fidelity is not the same as Pleasant Music
If a speaker reproduces some music with 100% accuracy and the result is unpleasant, doesn’t that just mean the original music—as created by the artist—is unpleasant?
Where possible, I’d prefer a speaker that respects the artist’s decisions instead of inserting itself into the creative process.
Unless you are listening through the same studio monitors in the same room or headphones as the mixing engineer, it will never be the same.
IMHO, people place too much importance on "accuracy". While accuracy might be objectively measured, it means nothing when it comes to individual taste.
There’s a whole field of research on this (look up Floyd Toole) - while any one individual can have skewed taste, on aggregate people prefer speakers that are as close to neutral as possible.
Signal reproduction matters quite a bit more for music production than it does for music listening and enjoyment. That's why producers and engineers look for 'monitors', rather than hi-fi speakers.
Hi-fi speakers, tube amps, and other accessories generally "degrade" the sound with added harmonics and natural smile EQs. That's what makes them sound more pleasing.
(I'm not disagreeing with you, just adding more color.)
> Great story. It is interesting how being stranded in some place with a computer and some skills always results in the most fantastic projects.
TLDR: IME, solitude is required for clear thinking.
-----------------------------------
Long ago, I used to drive 600km (one-way) twice a month . Kept it up for 4 years or so. As I drive with the radio off, I had much time alone with my thoughts.
Now, I wonder if always reading is having a negative effect: we're constantly bombarded with content all the time, and even though I never doomscroll (no tiktok account, no FB account, no instagram, etc), I think sometimes that enforced solitude might do wonders for my problem-solving.
I wonder how people who are on all those social networks ever find time to just ruminate.
Heh, that's a very timely comment. I just drove up and down to Berlin through absolutely crap weather and still figured something out I'd been struggling with for weeks.
The problem is that that's too many characters to reserve (they all have to be escaped when you want the actual character) making the resulting text look awful in plain text mode.
reply