ryanobjc

The multimaterial is going to be really interesting, combining TPU + PLA for example. Or TPU + PETG, who knows!

In addition to standard multi-color needs!

Prusa are clearly tools: you can fix it, modify it, and still have warranty apply. You can get after-sales parts, and service.

Bambu are appliances. They can work great out of the box, but appliances do not have upgrade paths. You do not upgrade a microwave, you throw it out and get a new one. Or maybe it's more like a fridge, you can limitedly repair some bits, but you cannot wholesale upgrade from V1 to V2.

Anyways bought the core one a few months ago, on kit, and did the whole assembly. The assembly was fun, and the resulting printer has been great. The print fails I had were all easily understandable, entirely due to adhesion issues to a dirty print plate.

I also ordered the indx and am looking forward to capabilities that are not possible with the AMS system. I'm more of an artist though, so I'm looking for interesting and cool ways to make things, not just 3D printing figurines or sculptures etc.

What value to a person does teaching "how to use it effectively" deliver?

How does that benefit their development, learning, society as a whole?

Before you start in with "it'll help them get a job", full stop - education as a public good isn't strictly vocational technician work. It's not a work training for companies.

I think "working product" is doing a lot of heavy lifting for you here.

I think anyone who has done serious product development wouldn't be so flippant and dismissive about the difficulties of even conveying WHAT to build, let along getting the right balance of quality, timeliness, and actual functionality.

The price hikes are going to be absolutely devastating.

Imagine oracle level price acuity along with 0 competition and utter dependence. This is the future the AI labs are drooling for. You will be charged based on the value it delivers. People will be start making trade offs on if hiring humans would be cheaper than AI, etc.

There's no way they're going to leave all that money on the table when there is all that investment to pay back.

That's the thing, often when people say stuff like "its banned" what they really mean is:

- the cost of mitigating the human health risk is too high - competitors in low-environmental regulation places don't pay for those costs - ongoing verification is expensive

I mean, let's face it, "self-regulation" of industries isn't really working that great. And for things that are health hazards that are basically borne by someone else, why should a local government make it easy to cheat and lie about this stuff?

The people arguing against this seem to assume that their right to have a business, make a profit, whatever, is a self-evident Good Thing, and rarely provide any additional arguments beyond "but the jobs". If they were at the VERY LEAST saying "we can make X safe" then maybe it'd be interesting. But as it is, the argument is basically asking us to mortgage the health and safety.

Paint VOCs sounds fine, until it's done at industrial scale, and it's also your neighbor, and also all the children in the neighborhood have asthma, and also healthcare is a lot more expensive...

This list isn't things you "cant do in california" but "polluting things you can't do in highly populated cities".

I'm not sure what the conclusion here is other than health is not important.

I've used chatgpt-shell, but I have since turned my LLM usage to gptel inside org-mode buffers. Every day I use org-roam-dailies-goto-today to make a new file and turn on gptel (the use of org-roam-dailies is 100% optional). Then I do my interactions with gptel in here, using top level bullets and setting topics to limit context.

I have 10 months of chats, and now I can analyze them. I even had claude code write me up a program do that: https://github.com/ryanobjc/dailies-analyzer - the use of gptel-mode allows me to know which parts of the file are LLM output and which I typed in, via a header in the file.

Keeping your own data as plain text has huge benefits. Having all my chats persistent is good. It's all private. I could even store these chats into a file.gpg and emacs will auto encrypt-decrypt it. Gptel and the LLM only gets the text straight out of emacs, and knows nothing about the encryption.

I found this better than the 'shell' type packages, since they don't always keep context, and are ultimately less flexible than a file as an interaction buffer. I described how I have this set up here: https://gist.github.com/ryanobjc/39a082563a39ba0ef9ceda40409...

All of this setup is 100% portable across every LLM backend gptel supports, which is basically all of them, including local models. With local models I could have a fully private and offline AI experience, which quality based on how much model I can run.

I mean... explain sora.

If I wanted to do that, I'd just move into engineering management and work with something less temperamental and predictable - humans.

I'd at least be more likely to get a boost in impact and ability to affect decision making, maybe.

The "just generate go code automatically then check it in" is a massive miswart from the language, and makes perfect sense because that pathological pattern is central to how google3 works.

A ton of google3 is generated, like output from javascript compilers, protobuf serialization/deserialization code, python/C++ wrappers, etc.

So its an established Google standard, which has tons of help from their CI/CD systems.

For everyone else, keeping checked-in auto-generated code is a continuous toil and maintenance burden. The Google go developers don't see it that way of course, because they are biased due to their google3 experience. Ditto monorepos. Ditto centralized package authorities for even private modules (my least fave feature of Go).

It's kind of weird how you downplay tardive dyskinesia, as if it was kind of a no big deal, whatever kind of thing. Would you accept having tardive dyskinesia induced by a drug?

Is that really your position?

Your other words seem fine, but that is a standout sentence!

regarding #2: "Automate the dumb/boring stuff", I always think of the big short when Michael Burry said "yes I read all the boring spreadsheets, and I now have a contrary position". And ended up being RIGHT.

For example, I believe writing unit tests is way too important to be fully relegated to the most junior devs, or even LLM generation! In other fields, "test engineer" is an incredibly prestigious position to have, for example "lead test engineer, Space X/ Nasa/etc" -- that ain't a slouch job, you are literally responsible for some of the most important validation and engineering work done at the company.

So I do question the notion that we can offload the "simple" stuff and just move on with life. It hasn't really fully worked well in all fields, for example have we really outsourced the boring stuff like manufacturing and made things way better? The best companies making the best things do typically vertically integrate.

The author is pointing out that aggregate productivity hasn't really gone up. The graphs are fairly compelling.

There are many reasons for your experience, and I am glad you are having them! That's great!

But the fact remains, overall we aren't seeing an exponential or even step function in how much software is being delivered!

Apparently Mac sales are seeing a big growth spurt lately. This isnt obvious at the Apple bottom line because iPhone dominates, but they are growing that market segment!

The apple silicon has driven a ton of people to laptops, and the windows 11 migration nightmare has given people an option: why bother buying a new pc laptop to run windows 11, just switch to the apple macbook air.

There was a quote recently where Nadella was trash talking the apple AI stuff, and like yes sure, but also do we not realize that Nadella isn't a random third party CEO, he's literally financially incentivized to maximize AI usage because Microsoft is selling the stuff?

It's like listening to the Intel CEO saying that the new apple silicon is alright, but what would REALLY make them go is to migrate back to Intel x86.

I'm not really upset at Nadella for this, because it's what he should do. But for everyone else to breathlessly be obsessed with his word and forget his real job as salesman in chief is what's pathetic.

I think the reason why Apple hasnt come out with such a useful personal assistant AI is because the underlying technology just doesn't make this so easy.

Apple is the only company so far that seems to be unwilling to accept a poor user experience with generative AI, so their efforts have been "lackluster" in terms of "AI integration" - as if "maximally integrating generative LLMs" is the goal in and of itself!

Of course Nadella is critical of Apple's efforts: he selling the goods! It's like a Steel CEO complementing a new bridge but then "they should have used more steel" - well duh, but since when do we trust the purveyors of components as to what should be added?

The bottom line is "Agentic AI" is just unreliable. If you thought you hated Apple Intelligence now, then if they had gone whole hog, the unreliability would be astounding.

As a long time Mac and iPhone user, I just don't get all the complaints about innovation. Macs are as innovative as they ever have been, maybe more. Apple silicon has been an absolute powerhouse and godsend for battery life.

iPhones are great, and continue to be awesome. Airpods are so good they make bluetooth headphones look like the garbage they are.

When you look at the grand scope of thing, the primary thing the commenters are missing are the Jobs' pageantry and showmanship. Which I also miss. But in terms of capability, I am quite happy with what we have.

From a CEO pov this move makes no sense, as evidenced by the fact that literally no on else does it.

It goes against all MBA orthodoxy.

I think there are parallels with functional languages on the JVM. The parts that are the worst are the parts that were built for maximum interoperability. Not to mention that the JVM forces classes on you at the deepest opcode levels.

Compatibility with C++ is fine, but so far it seems carbon's safety story is entirely a wishlist rather than anything yet. Seems like Carbon might be a more of a place to demonstrate features for C++ committees than a real language?

Personally I have hand it up to here with lousy programmingn languages that make it easy for me to write bugs.

Because there's a strategic benefit and the cost is practically negligible compared to the cost of this section of the economy going away.

That is the political calculation, not "throw good money after bad" kind of economics 101.

Just cancelled my Cursor sub due to claude code, so heavily agree.

So this has already been conceived of many decades ago, and there are some substantial issues with it, the illustrious djikstra covers it: https://www.cs.utexas.edu/~EWD/transcriptions/EWD06xx/EWD667...

Now this isn’t to say the current programming languages are good, they are generally not. They don’t offer good abstraction powers, typically. You pay for this in extra lines of code.

But having to restate everything in English, then hoping that the LLM will fill in enough details, then iterating until you can close the gaps, well it doesn’t seem super efficient to me. You either cede control to what the LLM guesses or you spend a lot of natural language.

Certainly in a language with great abstractions you’d be fine already.

This article is literally the definition of TL;DR. It's fairly hard to get thru, I spaced and skipped the conclusion, that the treatment isn't permanent and can be undone.

The notion that "many localities completely controlled by Democrats" is just not really true. The idea that zoning reform is a no-brainer and easy is also clearly not true: it's been a major teeth pull in SF and California, and many different people are pushing back.

Additionally somewhere like SF isn't a monolith, and despite voting heavily for federal democrats, the local elections aren't so clear cut.

Also we need to consider how and if development is truly good for existing residents. Many of them do not think so. The appropriate response here is to develop a sense of curiosity: why is that? What do people believe, and why?

I would say this: modern development doesn't always make the kinds of mixed used neighborhoods that people truly want and love. The cost of development often results in street level vacant commercial space, and that is not good for neighborhoods. Yes a complete revamp of zoning could change this - but would it? How do you build a compelling future that brings people along?

It just isn't trivial sadly.

A restaurant out of your home presents a public health risk sadly. Foodborne illnesses are rare, because of the regulations placed on commercial food sellers.

There are likely ways we can give more latitude, but home kitchens often cannot be cleaned to the same standard as commercial kitchens just due to how they have been built.

Unfortunately in terms of "overreaching" regulations, this is the worst example: food borne illneses are very real killers, and a major risk.

Also regarding a licensed electrician, while the regulatory requirements for one may seem high - I'm not 100% sure if they are or not - electrical fires are one of the major ways structures catch fire and kill people.

Really in terms of overbearing regulations, these two actually protect lives every single day. Remember the "Ghost Ship" fire? Caused by electrical fire. Norovirus on cruise ships? Cause hospitalization and evacuation.

I'm all for minimizing regulations, but many many of them are literally written in blood, and the notion that we can wholesale relax them with no ill consequence is just not true.

"If a school enroll someone with a 400 point lower sat over the higher person on the sole basis of their race, that is a major Injustice on the scale of individual humans, even if it moves some group statistic closer to equal."

So this is a hypothetical that is not worth discussing. Until specific cases can be brought to bear, why are you inventing situations that may never have existed?

I'd also like to opine on SAT scores for a second. First off, it's well known that SAT scores are not directly correlative in post-secondary educational success, nor work-success. Second they are not highly accurate measurements - there's an inherent fuzziness to them. So even if 2 students had a SAT within some Epsilon, the SAT scores might not really provide much differentiation there. Ergo, basing all of our policies on SAT scores - which are well known to be easily gamed, and also a product of a private institution - seems not a good idea.

Moving on, the problem with being against "myopic obsession with the group statistics" is you are ignoring some important evidence. What do you think of the "group statistics" that say that black people start less businesses, have less family wealth? Or black women have higher maternal mortality? These are pointing to important individual outcomes that are, to say the least, wrong.

So I don't think that paying attention to group statistics, like black maternal mortality (aka how many black moms die in child birth or due to child birth) is "myopic" and "to the detriment of individual equality." It's a very very real problem we, if we intend to call ourselves a moral society, need to solve. So having specific programs to help solve black maternal mortality in a hospital is not "countering racism with racism" imo. It's a focused program on solving a focused problem.

This logic extends out to most "DEI" things. For example is it good if the students at universities drift from representing America on average? I'd say it is not good. What about the ivy league pledges to make school free for anyone who's family income was under $X a year? Is that a myopic obsession with group statistics, namely poor people who can't afford elite colleges even if they were admitted? Seems like yes that could fit into your definition of why you oppose DEI. And IT IS a DEI program - it's increasing diversity (income/class diversity) and equality/equity (improving outcomes for individuals) and inclusions (including those who cannot afford elite colleges).

So when DEI programs that are focused on race, because much of our racial divide was artificially constructed by racist laws and policies of the past, it is suddenly bad, even though I rarely hear anti-DEI people go on about the low income scholarships for ivy leagues. Honestly it starts to sound that in fact many people may in fact have a problem not with the overall concept, but the beneficiaries of the programs.

So back to your comment, let's pick some specific circumstances that we know about and may you can propose how you'd meaningfully fix it, policy wise, within 5 years: - Black maternal mortality us 3x higher than white maternal mortality - Black people are ~ 14.4% of USA, but 12.5% in colleges. Is this a problem? - If we think talent is spread equally, then we should expect to see more % of black founders in YCombinator than the 2-4% there is. Maybe not exactly 14.4% but surely closer to 10% than 0%? Is this worthwhile of being solved? - While we are at it, only 11% of YCombinator founders are women. Is this a problem?

So what can be done about these noticeable gaps? What kinds of suboptimal outcomes are being picked when, for example, few YCombinator founders even know about the challenges and struggles of the average American? (who's a woman btw, women are 50.49% of the population, a majority) What kinds of products, opportunities, etc are being missed here? Maybe none?

What are your "race blind improvements" to economic mobility here? You have a 5 year timeline to make statistically meaningful changes to these metrics.

It's rather ironic since we know that women, poc, and more often face a lot of professional resistance, and therefore have to be better than average to succeed.

Which means when you come across a black or female professional who has risen, it means they actually are much more likely to be MORE talented than the average white man.

In other words, this notion of "diversity hires" is not logical. It barely makes sense.

So you said "make it easier for someone without rich parents to start a business by lowering barriers to entry and regulatory overhead on small entities"

This is a supposition: the cure "lowering barriers, regulatory overhead" may not cause the intended outcome "make it easier for someone without rich parents to start a business".

Given the primary reason why it's hard to start a business is access to capital, I'm not really sure what "lowering barriers" (which barriers exactly? how?) and "regulatory overhead" (which ones specifically?) will meaningfully do to improve the outcomes of black people.

And this is before we even talk about the well documented facts of biases, outright racism, and uneven application of laws.

So, how do we get to the outcome we all want: your talent drives your success?

One way you could do this is to have government programs to provide startup capital to certain groups. You know, like we already had, but are attempted to being erased under the "anti-DEI" crusaders.

In reality a lot of the anti-DEI rhetoric is based on disinformation, misinformation, and honestly just good old fashioned racism.

It’s not often you get 100x optimization with some small improvements so I’m kind of skeptical.

We have and apples and oranges thing here which deepseek is intentionally leaning into. They get very cheap electricity and are bragging about their cheap cost, and OpenAI etc typically brag about how expensive their training is. But it’s all pr and lies.