I ran OpenCode + GLM-5.1 for three weeks during my vacation. It’s okay. It thinks a lot more to get to a similar result as Claude. So it’s slower. It’s congested during peak hours. It has quirks as the context gets close to full.
But if you’re stuck with no better model, it’s better than local models and no models.
I have to say, OpenCode’s OpenUI has taught me what modern TUIs can be like. Claude’s TUI feels more like it’s been grown than designed. I’m playing around with TUI widgets trying to recreate and improve that experience
> I have to say, OpenCode’s OpenUI has taught me what modern TUIs can be like. Claude’s TUI feels more like it’s been grown than designed.
Claude's TUI is not a TUI. It's the most WTF thing ever: the TUI is actually a GUI. A headless browser shipped the TUI that, in real-time, renders the entire screen, scrolls to the bottom, and converts that to text mode. There are several serious issues and I'll mention two that do utterly piss me off...
1. Insane "jumping" around where the text "scrolls back" then scrolls back down to your prompt: at this point, seen the crazy hack that TUI is, if you tell me the text jumping around in the TUI is because they're simulating mouse clicks on the scrollbar I would't be surprised. If I'm not mistaken we've seen people "fixing" this by patching other programs (tmux ?).
2. What you see in the TUI is not the output of the model. That is, to me, the most insane of it all. They're literally changing characters between their headlessly rendered GUI and the TUI.
> Claude’s TUI feels more like it’s been grown than designed.
"grown" or "hacked" are way too nice words for the monstrosity that Claude's TUI is.
Codex is described as a: "Lightweight coding agent that runs in your terminal". It's 95%+ Rust code. I wonder if the "lightweight" is a stab at the monstrosity that Claude's TUI is.
This might be the worst idea I've ever seen. I'm glad they are so interested in reducing the effects of global warming - that's fantastic - but they are literally purposefully releasing a toxic, major air pollutant into the air to create in their own words "clouds of dust" for the purposes of reflecting sunlight. Sure, there might be a slight cooling effect but who in their right mind could possibly think this is a good idea?!
I understand they are deploying to the stratosphere and not the troposphere but I can't imagine there aren't any negative second-order effects.
As someone who lives in a city with a major PM2.5 problem that effects the millions here on the daily (near an active stratovolcano no less!), reading about what they are doing was somewhat infuriating.
I appreciate the link, however it should be noted that piece was written by the co-founder of that same company. They do seem to be evidence driven but let's not pretend there's not some degree of bias towards the solutions his company is proposing.
Additionally, that article focuses almost solely on the chances of producing acid rain, which actually is another issue and not the one I was first concerned with. That piece talks about "redistributing" SO2 from the troposphere to the stratosphere which is a neat concept but a.) that's purely theoretical, and b.) that's not what they are doing or trying to do anyway.
It could be argued that air pollution has a greater and more devastating effect on the everyday lives of people alive today then global warming does now or will in the foreseeable future. In my city it was estimated that more than 1 in every ~16 deaths is related to air pollution, and the air here isn't nearly as bad as it is in other cities. Worldwide, UNICEF estimates nearly 2,000 deaths under 5 years of age per day from air pollution [0]. Annual deaths worldwide are estimated at 8+ million yearly across all age groups. Making that problem worse by any measure in the hope of producing a cooling effect that is a fraction of a fraction of a degree is not worth it at all and at least in my opinion is a net-negative.
English can be read in a different order than the normal order when the sentences contain words for which it is easy to guess whether they are agents or patients, e.g. when the agents are animate nouns and the patients are inanimate nouns, or when pronouns are used for the agents or patients.
Otherwise, the non-standard order can be understood incorrectly. While the distinction between agents and patients is the most important that depends on word order in English, there are also other order-dependent distinctions, e.g. between beneficiary and patient, when the beneficiary is not marked by a preposition, or between a noun and its attribute, e.g. "police dog" is not the same as "dog police" and unless there is a detailed context you cannot know what is meant when the word order is wrong.
English is one of the languages with the most rigid word order. There are languages, especially among older languages, where almost any word order can be used without causing ambiguities, because all the possible roles of the words are marked by prepositions, postpositions or affixes (or sometimes by accentuation shifts).
The deterministic part (calculations) is done by Excel.
The non-deterministic part is turning human instructions ("calculate the NPV over 10 years for X given Y") into Excel.
This is already a non-deterministic process (humans are non-deterministic!). The question is if an AI model can be more reliable than humans, and I can't see any reason why it wouldn't be.
The correct path is pretty clear, so the logits for following that path are going to be a long way from off-path.
For something like this the real problem is training the model to use Excel (which will show up by it being confused which sheet it is on or trying to use the wrong window or things like that), not the non-determinism.
so basically what you're saying is that: it doesn't do the math, it tells the math-doing-thing what math to do. Basically, instead of humans using Excel, imagine AI using Excel?
Yet I don't understand the aha moment here? It might save analyst time but aren't there already enough automation that you don't really need to tell the AI to tell the math-doing-thing to do the math because the math-doing-thing is already optimized for most general functions? What are we gaining from adding the non-deterministic process here when the real non-deterministic process is still the human being prompting what to do?
Seems like a solution to a non-problem from my pov.
Sure in the sense that you're setting up a program that has inputs and outputs etc. But then all math is programming. All language is too, even speaking can be considered programming if you're stretching the definition enough. But I will disagree that setting up spreadsheets = basically software engineering.
The LLM just needs to make sure it uses it appropriately. Doing that bit is the non-deterministic part, but the NPV calculation itself is completely deterministic.
When you setup a spreadsheet, you choose a column or cell to show the NPV. You use the npv formula on this cell and apply it to inputs from other columns or cells.
This process of deciding what data to put in what column and how to apply NPV to it is non-deterministic. You could choose to put it in column A, or column B or maybe even row 3 - it depends!
Do you really think you come off as a good person posting things like this? This place is to satisfy intellectual curiosity, not to lord your supposed intelligence over others, without even proving it exists in the first place.
I read something at some point that it's more expensive to convert these into residential buildings than it is to literally demolish and rebuild.
I'm not entirely sure how that math works out, or why, because one would think it couldn't be that complicated. Maybe someone here knows more about this.
Another thing about a lot of commercial buildings is the floorplate size and layout. Office buildings often don't care if there's a lot of interior spaces without any windows, but people need outside light. So if you've got a massive floorplate it can be kind of a pain chopping it up into good sized units that meet the demand of the residential market in the area. This definitely varies from building to building though.
There's also a lot of work that probably needs to go in to the ventilation and fire code changes. An office building isn't designed for people having ovens and stoves. It also often just assumes its OK to have less isolation between units for the ventilation, or previously entire floors were considered to be one space ventilation-wise but now you might be trying to split it into 2-3 units that require separation. This separation can also complicate things like AC and heat.
The ventilation issue comes up a good bit with a lot of these poorly done conversions. You end up with units that just don't get nearly enough airflow, and all the windows are sealed so its not like one can just open the window to get more air.
The plumbing systems in commercial buildings are not big enough to handle residents usage. Residents use more water and the outbound sewage systems need to be larger.
There's already enough plumbing in there for a whole office to shit when they get to the office.
History favors the bold, and code inspectors blabbering about "written in blood" don't see all the homeless people they kill via reduced access to housing.
I've seen plenty of artist collectives that manage it; on paper they are office/industrial but actually everyone lives there. Every once in awhile one burns down but the mortality rate isn't as high as living on the streets which is ultimately what happens to those on the bottom of the socio-economic pyramid when the ones higher up push the ones under them down a rung to snag housing.
For a lot of the office buildings I've been in, there aren't that many toilets per floor. Its also different when you've got some toilets that are often unused compared to people running laundry, cooking, bathing, etc. Very different demands on the plumbing system.
You also then had everything pretty much isolated to two rooms for an entire floor meanwhile now every unit is going to have a separate kitchen, a bathroom (or two, or three), a laundry room, etc.
And you're going to need a good bit of engineering studies done before you start cutting that many holes in the floor.
Ok, but some extra plumbing (and whatever sorts of engineering studies referred to) and electrical work surely can't as expensive as demolishing and rebuilding a whole building.
If it was just the plumbing, then maybe. But its not just the plumbing. Its the plumbing, the electrical, the AC/ventilation, fire codes, and so much more.
Not saying it can't ever be done, it really depends on the building. But its not necessarily a good assumption it can be done well in a cost-effective fashion.
But do you really have to cram in as many residents as you could with a purpose-built tenement? There must be ways to keep headcount in the range the infrastructure can support and still provide a lot more housing than just leaving them as empty decaying offices owned by the last one holding the bag. Intersperse flats with windowless storage units (you have a depth problem anyways), low density commercial use like workshops with live-by flats and so on. Large units designed to attract high squarefeet/low headcount tenants, not bunk bed families. Add regulation only as a fallback limiter. Don't let perfect be the enemy of good.
> There must be ways to keep headcount in the range the infrastructure can support and still provide a lot more housing than just leaving them as empty decaying offices owned by the last one holding the bag.
Demolishing the office building and building a residential building is more profitable often.
I mean, sure, you can just sell it as a unit for each floor. You then need to recoup all the costs of rebuilding against fewer people, so all the main area renovations and what not get more expensive and the monthly cost of building maintenance get spread across fewer and fewer tenants. But you've still got a problem of most of the rooms of your very expensive condo have zero natural light, its all practically ancient built stuff in terms of planned structure life, and you've got a very expensive monthly maintenance bill. Meanwhile your massive and dark unit with odd plumbing and low ceilings is competing in the market against units that were actually built for the purpose of people living in them, so while your unit is big and expensive to maintain they're some of the least desirable spots.
The economics just often work out a lot better to tear down the old structure and rebuild a new one more fit for purpose.
Sorry, I either totally misread your comment or was mentally replying to someone else when I wrote this.
Sure, you could just cram the residences to the edges and try to recoup the cost of the rest of the square footage for places that don't need natural light. But once again you've got issues with original designs and intents for the building. None of the plumbing is designed to be pushed to the edges, so you'll need to make massive changes to the structural integrity by drilling a bunch of new floor cores to do all the new plumbing work. You could rent the interior spaces as storage, but you'll probably quickly flood the market of storage units with the massive amount of square footage you'll be bringing.
Trying to have industrial in there as well is asking for problems. Trying to rent some 15th story small/medium interior unit as some kind of industrial workshop would be quite weird. What kind of industry would want a smaller interior space that probably can't support heavy equipment, has a limit to ceilings of ~10 or so feet, can't require odd ventilation or strange/additional fire suppression/separation requirements, probably has significant power limitations (in terms of industrial capacity, at least), noise limitations, difficulty getting much product in and out, etc? Stuff that the city is going to be OK zoning literally across the hall from people trying to live? And that you're going to find a number of these willing to pay a good bit for such a space to cover the maintenance costs? These buildings weren't built for industrial usages, they were built for office desks and couches. Maybe a few floors have been upgraded to handle additional weight to have datacenter kind of spaces, but definitely not most of the floors.
So then you're trying to spread the maintenance costs of this massive and old building across higher value residences and a lot of very low value storage/weird industrial tenants.
You can run drains out the side of the structure without drilling holes in the floor, same with electric, and even if by some insanity we say "whutabout the holes in the side" then you could even use a damn lift pump/macerator pump to pump it up and out through where a window was. For vents you can also use AAV instead of a traditional vent. If the residences are at the edges they should be able to pop right out and worse case you elevate the floor in the bathroom/kitchen under the plumbing appliances for the slope on the pipe as it exits. A vertical drain pipe isn't going to freeze (and even if it were, could be insulated and heated), and supply lines are such small holes as to not threaten structural integrity.
> And you're going to need a good bit of engineering studies done before you start cutting that many holes in the floor.
You can Swiss-cheese a pan and deck concrete floor with core-drilled holes, the important thing is GPDR scanning before coring to avoid the pre- or post-tension cables embedded in the concrete.
Artists are a shrinking population, I wonder if having most of the top floors (20 out of 30) converted to extremely large luxury apartments (5000sqft+) and only 'adding capacity'to plumbing and what not for the lower 10 floors, which would house smaller units, would be economically viable. Although actual luxury market requires high ceiling so probably wouldn't work out.
I'm sure many many people have thought of all sort of solutions as the value for finding some sort of solution is extremely high.
> There's already enough plumbing in there for a whole office to shit when they get to the office.
A 20,000 sq ft office tower floor will usually have a single set of restrooms and a couple of kitchen sinks, maybe a dishwasher, plus a couple 6-gallon or instahot water heaters. If you subdivide that floor into a dozen units, that’s 12 showers, 12 washers, 12 dishwashers, 12 toilets, 24 sinks, and 12 water heaters.
The riser and drain pipes aren’t big enough to handle residential needs.
That's not how the 'black market' ones I've seen operate. And I've seen a lot from when I visited the circuit of underground artist-related events when I lived in chicago. They are shockingly common in areas with extremely high rents and an oversupply of unused commercial space.
They might subdivide it 12 ways, but there is one shared kitchen for a whole floor and maybe 2 toilets, 2 sinks and the residents are going to the laundromats. They tend to put the shared amenities on the ground floor as much as possible because it is easiest to expand them there. It beats being homeless by a long shot.
For reference, when I hauled water, we used about 60 gallons a week for a family, or about 0.05% utilization of a 3" drain pipe for a single family. You do not need much water in order to be way way better off than being homeless; 5/gal a day of non-potable water and you're pretty much in luxury comparatively and a shit-ton of people can be putting that down a 3" or even 2" drain pipe before it causes problems. A 3" pipe is the minimum that would be serving a typical floor of a warehouse, so plenty enough for a constantly used couple of shared bathrooms with a shared kitchen. Honestly even splitting it 12 ways could be overcome with some technical ingenuity (electric lock-outs to prevent more than a few in use at once, and AAVs to prevent needing a bunch of new vents).
These are all easily overcome problems for people utilizing an ounce of civil disobedience with regards to the code. And yes I have personally done all the design and plumbing and electric for multi-structure properties (though not the black market ones).
> These are all easily overcome problems for people utilizing an ounce of civil disobedience with regards to the code. And yes I have personally done all the design and plumbing and electric for multi-structure properties (though not the black market ones).
It didn’t work out so great in Oakland at the Ghost Ship, 36 people died in a similar arrangement.
Building code is written in blood, things are done a certain way for a reason. You may be morally or ethically against them but following code saves lives.
36 people dead is a rounding error compared to mortality from people on the streets due to lack of access to housing. Every time I bring up this topic, someone trots out the Ghost Ship like a broken record, ignoring what I said about the mortality rate of people on the streets because shit rolls downhill when people higher up the socio-economic pyramid go the next rung down in available housing. Bastiat has an excellent writing on this fallacious logic you use, titled "That Which is Seen, and That Which is Not Seen."
Not having housing didn't work out great for 700+ dead homeless people per year that are estimated to die of hypothermia.
The code inspectors have blood in their hands. You may be morally or ethically against bypassing the codes, but bypassing it can save lives.
Black market housing is done for a reason, a very good one, and one that saves lives. Fortunately where I live, I built a house without any inspections whatsoever, so none of the code psychopaths were even around to make their absurd case about the ghost ship, and that is the only reason why I was even able to afford to own a house.
This is an issue that got brought up in Portland, OR during Covid IIRC. The city was looking at buying up vacant offices and converting them to living space but it just didn't make any sense financially and the city concluded it was cheaper to demolish and rebuild than convert.
I assume it's because they would need to re-wire electrical and retrofit plumbing on a massive scale to accommodate kitchens and bathrooms for separate units. They end up needing to gut the entire building and cut through floors and ceilings without damaging any structural and load-bearing parts. It doesn't sound easy nor cheap.
>at some point that it's more expensive to convert these into residential buildings than it is to literally demolish and rebuild.
Yep, and that's fine. It's literally a tangible instance of 'creative destruction'. I see people arguing that oh, we have to RTO to save the current model and it seems so backwards to me.
My team is doing the same, and yet all of us still aren't sure that we're actually more productive overall.
If anything it seems to me like we've just swapped coding with what is effectively a lot more code review (of whatever the LLM spits out), at the cost of also losing that long term understanding of a block of code that actually comes from writing it yourself (let's not pretend that a reviewer has the same depth of understanding of a piece of code as an author).
If you work in a team then you are likely already not writing most of the code yourself.
There will be point where ai will consistently write better prs - you can already start to see it here and there - finding and fixing bugs in existing code, refactoring, writing tests, writing and updating documentation and prototyping are some examples of areas where it often surpasses human contribution.
All the comments of AI writing code and making PRs remind me a lot of all the promises about self driving cars. That was more than 10 years ago and today I still don't know anybody that has a car that drives itself. Will AI write useful PRs one day? Probably. Will it do that before I retire or die of old age? Considering I have been using agents for about a year or so, and seen little to no improvement in that time, I'm afraid the current version of AI probably has already peaked and we'll only see marginal improvement due to diminishing returns.
Yes there is a very real trade off between labour and capital.
In the past the tradeoff has been very straight forward. But this is a unique situation because it involves knowledge and not just the physicality of the human in regards to productivity.
React + Vite + Tailwind on the frontend; Netlify Functions for backend with Stripe, Supabase, and email integrations; content via Markdown build script; deployed on Netlify; linted with ESLint; JavaScript-only codebase
The product label images loading on the homepage are huge right now. They are displayed in 128px * 128px box but are about 2 MB in size each. May be generate resized versions at build time and use <picture> tags?
reply