domain:mattlakeman.org
Building more transit is doable. Making transit good is not.
You know this has lead me to an important realization. If it came down to it, I would choose to be ruled by the wokest HR lady on Earth before this guy. As much as wokeness disgusts me it doesn’t really scare me, it doesn’t strike me as outright evil or insane. I’m switching teams as of now.
Fuck, I’m sorry.
I have stress-provoked tachycardia which is less crippling but certainly unfun and produces perhaps a similar ‘I need to do important things to have a good life but not too many things or else my body will go wonky’ dilemma.
I’m sort of failing to manage it at the moment but generally I would advise trying to separate out one’s stressors a bit. Don’t go drinking the night before the marathon. Try to have specific study times that don’t coincide with your worst work shifts. If possible, try to have the confidence that you can take a break every so often without fucking up your life.
As I have lamented before, I barely play video games these days.
True, and I apologise for not being around to play something as I said I might be. Work picked up a bit but mostly I’m too lazy even to be properly thoughtfully lazy…
I have my doubts about getting a massive screen because of eye strain. I’m told that if you have a big screen dead in front of you (ie you’re looking straight ahead rather than down a bit) your eyes instinctively try to focus on the horizon and it messes you up long term. I work on this thing most of the day so I want to be careful about that.
https://cluvens.com/scorpion-ergonomic-gaming-chair.html
A friend and I have a pact that when our careers takes off and we become immensely rich we will buy one of these lol
You have exams?
No, it isn't. I'm not talking about remembering a bunch of explicit instructions or preferences. I'm talking about learning in the way a competent person goes from a newbie to a domain expert. That is completely missing in LLMs. No matter how much I guide an LLM, that doesn't help it generalize that guidance because LLMs are static snapshots.
If you want truly online learning, you're in for an indefinite wait. Fortunately, most people get a great deal of mundane utility out of even static LLMs, and I'm not sure what you need that precludes this.
And if your answer is "but GPT-6 will totally have been trained better", then why on earth would I waste any time whatsoever with GPT-5?.
Because... it's the model we have? Can't have tomorrow's pie today, even if we're confident it's going to be tastier. Why buy an RTX 5090 when Nvidia will inevitably launch a better model after a few years? Why buy a car in the dealership today when you can wait for teleportation with complimentary blowjobs?
If you're trying to force everyone to use the solution you like, you better be damn sure your solution actually works for them instead of constantly resorting to "no, you're just using it wrong".
Hold your horses buddy. When have I forced anyone to do anything? @SubstantialFrivolity has clearly articulated his concerns about the weaknesses of LLMs as of Today AD. I invite you to tell me which of his concerns online learning is strictly needed to address? As far as I can tell, I have emphasized that his boss has a point, or is directionally correct, and that he could benefit from using LLMs more. I hope you've noticed multiple caveats and warnings attached.
If you are so convinced that even the best LLMs today are a waste of your precious time, then good luck with whatever you're using as an alternative. It's not like they're so entrenched that you can't lead a productive human life without one. They also happen to be very helpful for most people.
Yes they can
Source: I live in downtown Toronto very close to three grocery stores only marginally smaller than the suburban ones
Yes, if you want one outlet added or a p-trap replaced you should get a handyman to do it without a permit or do it yourself and just not tell the city. I just said that. But if you need a big job then we do live in a society. Some of the rules are stupid(setbacks etc) but lots of big jobs don't impact them.
Where on earth do you live that's somehow walkable but also has literally exactly one grocery store?
Walkability is enabled by density and density by definition means there's lots of stuff around, I am within 4 city blocks of 3 (soon 4) big box grocery stores, at least 3 fruit markets, and if you expand that radius to a ~15 minute walk you can add at least 2 more big box stores and 2 entire neighborhoods of places defined by their vibrant collection of grocers and other food stores.
The non-union contractors I've dealt with will add new plumbing and cut a vent hole in a ceiling. Specifically the same guy did both for me. Even the American ones quoted two prices: one higher price with permits and second lower price without permits. They aren't strict rule followers.
I literally live across the street from a large grocery store in the densest part of one of North America's biggest cities and the only time I go more than once a week is when I forget something, in which case, it's really nice it's right there...
I genuinely don't know what windmill you're tilting at right now
"You'll live in the pod, eat the bug, and only shop at cornerstores that don't have all the groceries you'd like to buy in one trip" isn't a WEF conspiracy lol
Nor can parking; transit is so bad
....so build more transit?
This all sounds like a problem of will and not an actual material problem.
I went blind again. Before anyone panics, it was for less than an hour, and I'm fine now.
I was previously diagnosed with a condition known as central serous chorio-retinopathy, where fluid leaks out of the vessels below the retina and makes it bulge out. It is usually due to corticosteroid consumption, which I've never done, but also due to prolonged and severe stress. Guess what I'm experiencing?
This isn't the first time. In keeping with a now-obvious pattern, it happened to me before a high-stakes exam. Once before the PLAB 1, again before the MSRA. And now, the MRCPsych Paper A looms ahead of me.
My stimulants might contribute (they're stimulating the sympathetic nervous system too), even if I take very reasonable doses. Unfortunately, my ADHD is not an affectation, I can't study without my meds, let alone pass exams. Especially exams that require months on end of grinding and memorization, when is rather be doing anything else.
So yeah, same choice as usual: lie flat, or keep fighting. I'm inclined to do the latter unless the attacks become so severe that I'm at great risk of permanent visual damage. I did see a doctor once, and it was decided that a waitful watching system was appropriate, instead of jumping to options like intravitreal injections or lasers. It's been a year and a half since then, and this attack was mild, so I suppose it wasn't the worst advice.
Anyone else have a few dangling Damocletian swords above them? Aimed at their eyeballs? Alternatively, what's the best way for me to manage my stress, when giving up or foregoing my meds isn't an option I'm willing to consider?
(I'm going to order myself some green tea. It helped in the past.)
I mean, what country can you point to where lots of citizens choose public transportation over automobiles for non-economic reasons?
NYC, Toronto, Japan, Germany, London, presumably some tier-1 Chinese cities like Shenzhen or Shanghai, Hong Kong/Singapore(maybe?)
Depends on whether souls, angels and demons end up coming from an existing theological practice.
Your premise is faulty. Most LLM front-ends have memory or instruction features. You can literally make sure it remembers your preferences and takes them into account by default.
No, it isn't. I'm not talking about remembering a bunch of explicit instructions or preferences. I'm talking about learning in the way a competent person goes from a newbie to a domain expert. That is completely missing in LLMs. No matter how much I guide an LLM, that doesn't help it generalize that guidance because LLMs are static snapshots. And if your answer is "but GPT-6 will totally have been trained better", then why on earth would I waste any time whatsoever with GPT-5?.
Like I said I have no use for or desire to be saddled with an annoying intern, whether a human or an LLM.
If you're trying to force everyone to use the solution you like, you better be damn sure your solution actually works for them instead of constantly resorting to "no, you're just using it wrong".
Some degenerate case where a cute town of 150k goes crazy building Connestoga hut villages and a million single people move in that are attracted by the $500/month rent
Traffic goes from easy to abysmal.
All public parks overrun with trash and dirtbags.
Average tax revenue per person craters so police and other services become unavailable.
People paying all of the taxes move away.
Town basically becomes a refugee camp.
The lawful system is supposed to be able to handle all the situations; after all, by definition, going outside it is forbidden. If it can't handle many common situations (and indeed getting a licensed contractor to do a small job is often simply not possible or feasible), it's broken, regardless of "we live in a society" defenses like "licenses exist for a reason".
Yes, if you get everyone to do everything they want and need to do within their little neighborhood, you can do that. Places like that in the US either tend to be planned retirement villages, or places which are extremely not-nice to live.
Uh.. Your premise is faulty. Most LLM front-ends have memory or instruction features. You can literally make sure it remembers your preferences and takes them into account by default.
My custom instructions on ChatGPT include:
Never do any calculations manually, make sure to always use your analysis tools or write a program to calculate it.
And guess what? GPT-5 is absolutely scrupulous about this. Even for trivial calculations, it'll write and execute a Python program.
I, or you, could easily add something like:
"Always use your search functionality to review factual information. Always provide citations and references."
A more sensible approach would be to let it exercise its judgement (5T is very sensible about such things), or to tell it to do so for high stakes information.
So, yeah. A non-issue. It's been an effectively solved problem for a long time. You can even enable a general summary of all your conversations as part of the hidden context in the personalization settings, so the AI knows your more abstract preferences, tendencies and needs. It's even turned on by default for paying users.
building more places to live so they're cheaper
Phrasing things in positive terms doesn't somehow make them positive. You could frame the building of Chicago's Cabrini-Green projects the same way.
Yeah, Austin is a shining example of how to deal with the problem well.
Tokyo is crowded and living there seems pretty cool
NYC is extremely crowded and I am strongly considering moving there
the shopping centers nearby are so crowded
Just build more shopping centers??? Seems like an easy solution
What bothers me about angry "I hate that things changed" posts like this is that it's based on a belief/argument that the status quo was fine, which it was not.
The status quo in this case is 1) ever worsening traffic as population scales via horizontal expansion, but road network capacity does not 2) a perpetual increase in housing prices causing the following (but not limited to): lower birth rates, higher homelessness + higher crime as a result, a general erosion of the Western social contract, lower economic growth from the friction of moving, higher property taxes due to less economies of scale, and more!
If you're going to oppose building, you need to propose a different solution to the status quo, which again, ISN'T WORKING
Trump Falls in Love with Me, a White House Janitor
Truly the humble H-doujin is the ultimate artform, primal neuron activation captured and distilled into pure animal spikes of dopamine. Perfect in its simplicity, universal in its application. Through dick, unity.
It's been described to me by a lutheran friend like this: ELCA are just autistic Episcopalians(with the variance, and age, that you'd expect), LCMS are conventionally conservative, WELS are so fundamentalist they rival tradcaths and quiverfulls.
That seems, broadly speaking, accurate.
More options
Context Copy link