D13Fd

There are a few different types of Markdown apps. Many of them work like Notational Velocity/nvAlt/Obsidian where it's a folder of plain text files, you type in Markdown primarily, and the editor just applies some formatting after the fact, akin to syntax highlighting. IMO its much nicer than looking at just the Markdown, but it also has the advantage of remaining 100% portable text.

I’m running 27B on a 5090 as well, and the results have been really strong. It does almost as well as, and sometimes better than, a 121gb DS4 model running on an M5 Max 128gb. 27B also flies on the 5090, and at medium think it returns results many times faster than my DS4 setup (the default xhigh is basically broken, though).

For the kinds of things I use a local model for (legal document review), it’s just spectacular. It also has good vision support. I’ve been using 27B more and more over DS4.

Why would an issue in the SpaceX data centers impact Claude / Anthropic? Work shifting over from Grok->Claude?

The example is particularly bad because in Skyrim the blacksmith actually does react when a dragon arrives (in one of the villages where the blacksmith is on an exterior cell) and can actually be hurt by the dragon.

I got what they meant by "OS" about 10 seconds after landing on the page, but I still think it's a bad term here. The system is not controlling the operation of the home or of the people within it. It's a home-centric to do and calendaring app.

I’m fine with Firefox and run it as my main browser on all my devices. But their corporate behavior means I’ll never give them my money again.

Excellent analogy. This paragraph invalidates the entire post and honestly just looks lazy. The author may be right anyway, but with that level of experience with these tools, he is really just guessing.

I’ve been running DeepSeek 4 Flash, Qwen 27B and Qwen 9B on local hardware. They work well for coding and document review tasks. I think Qwen 9B local on a 5090 might be legitimately helpful for small task agents in omp, since it is ridiculously fast. But my motivation is that I have data that I unfortunately can’t share with 3rd parties.

I have been eyeing a 512 GB Mac 5 Ultra to run full DS4 pro locally, which I expect would be pretty amazing as far as quality/recall. The only downside is that the speed is a lot slower than something like 27B on the 5090.

Because I felt the opposite a while back when I switch from OpenAI->Anthropic, and because I understood that the two companies were basically at parity right now on their frontier models.

Yeah but Sol shows it is possible to just send the readable explanation in the first instance. And I don't want to spend tokens and time on asking for a better version of each response.

Fable is much better than Opus 5 IMO but it just burns through tokens ungodly fast. I can hit my weekly Fable limit on a 20x Max account in a day.

I think you’re right. It’s not optimized for some kinds of work. My little project has a Textual TUI interface that needs to display a few hundred thousand rows in a table. It takes 14 seconds to load in the default datatable component. I instructed Opus 5 to replace the datatable component with a fasttable alternative, a new dependency. I let it go overnight.

When I got back up, it had spun for hours and proudly announced that, instead of doing that, it had optimized the datatable build and avoided the dependency, because the new datatable loaded in 11 seconds. Once I got it to actually make the fasttable version, it loaded in less than a second…

I’ve been doing some heavy work on a personal project lately. I burned through the limits on Claude, the plus a few hundred dollars in credits, and ultimately decided to move to an OpenAI account just so I can keep going.

I was surprised to find that OpenAI Sol is much much nicer to work with than Opus 5 or Fable at the moment. Especially on Opus 5, the way it communicates is just exhausting. It keeps “being honest” and “confessing” mistakes and just generally talking a lot. I felt like I had to really dig to see what it’s doing.

The project involves OCR, and despite repeated instructions not to, both Claude models keep spinning out a bunch of agents to re-invent the OCR setup, and they inevitably seem to invent a primitive serial version that takes 20x the time, or longer, to complete, and then running it against thousands of docs. Basically I have to watch it like a hawk or it just spins out on red-teaming tasks that take hours and hours.

I don’t know what its system prompt is, but Sol/Codex is just so much nicer to talk to. It only asks exactly what’s needed, it tells me only what I need to know, and it is just generally workmanlike. And it has not once decided to spawn an agent that spends hours pointlessly burning tokens and CPU cycles re-inventing the OCR process. I’m really liking it.

Is it, though? It was highly effective with basically zero downsides, and well executed. I’m honestly not sure what the problem is.

I have a keyboard I bought in 2020 called the Makerdiary M60 keyboard. It runs Python, and over the years I've coded in various functions. But it was somewhat buggy, and the upstream firmware was never upgraded.

I love the hardware because it's the only keyboard I've been able to find that keeps a live Bluetooth connection and lets you toggle back and forth between USB and Bluetooth instantly using only a keyboard shortcut. That means you use two computers, and switch back and forth as fast as you can type. Years ago I also added a feature that changes the backlight color based on the active connection, so you can instantly tell which computer your typing will go to. I haven't been able to find any other keyboard that's quite as nice.

Lately I've spent weeks using an LLM to update the extremely old keyboard firmware to the latest firmware of its components (e.g., Circuitpython, the BLE firmware, the boot loader) and implementing all of my dream keyboard features that I never had the gumption to spend hours implementing pre-LLM. I finally resolved years-old bugs, and I added things like layer-based backlighting (so that the individual keys with relevant functions are lit), a better battery gauge, better sleep/suspend timing, backlight calibration, better debounce, and so on.

I've also upgraded the hardware and added batteries to each of the keyboards (I have three of these, one at home, one in the office, and one for travel).

It has been a very satisfying project, finally scratching a years-old itch.

Exactly. And having AI to work out the 2am dependency issues actually makes maintenance a lot easier. This seems like an odd message.

I agree. I assumed an optimal path would leave the mower on the outside of the property, to easily return to storage. But the “optimal” solutions in the article leave the mower wedged between three rocks. That seems rather sub-optimal to me.

I’ve been using this for weeks, and it’s pretty incredible. I used to get motion sick especially quickly when I looked at my phone in a moving car. Now it’s the opposite, if I’m sick in a moving car, I try to look at my phone, which magically fixes it.

It’s amazing how that sick feeling in your stomach can be fixed by looking at some purely visual moving dots.

There is zero reason to shred 18th century books. Any such books are out of copyright.

This is the result of our copyright law in the United States, which is extremely tilted to favor authors and publishers. The judge made exactly the right call and the companies are following the law.

The fix here is to change the law to permit training AI without destroying the original materials. But that is going to be a heavy lift.

It’s a maxed out current-gen MacBook Pro w/ 128gb ram. I lucked out in getting it before they upped the price.

It runs really well and fast enough for what I need, and it works great, but it’s slower than Claude. I use it for projects where I’m not allowed to use third party AI for legal reasons.

I'm surprised that their home page basically acts as if LM Studio and others don't already do this. It's not clear what the difference is from a glance.

It also omits Open WebUI. I've been running Deepseek V4 Flash locally on my Macbook Pro for weeks using Open WebUI + DS4.

I agree completely. But in the end permanent daylight time is better than flip flopping every year, and much more achievable.

I think if people understood that daylight time = get up early and standard time = get up late, most would agree that getting up late is better.

I’m asking a question to determine his view of “conventional wisdom,” which could be all kinds of things. Why would you assume that his view of “conventional wisdom” necessarily includes cutting all caffeine?

Did you cut out all caffeine?

Yeah I still rewatch it occasionally. I feel like it’s full of missed opportunities, and you have to shut your brain off, but it’s still good fun.

This is so true. It’s much harder for drivers to see pedestrians these days. And walkers now have to deal with E-bikes and electric scooters flying down the sidewalks at high speed and silently. I’ve nearly been hit many times even as a very visible adult.

Six kids here, I use the “we needed some backups” joke all the time.

It's hard to even tell any more. So many people are using AI that even the non-AI-using people are starting to write like that.

> LLMs are useful for research and checking your work.

I have to disagree that it's good for LLMs to do the research, depending on the context.

If by "useful for research" you mean useful for tracking down sources that you, as the writer, digest and consider, then great.

If by "useful for research" you mean that it will fill in your citations for you, that's terrible. That sends a false signal to readers about the credibility of your work. It's critical that the author read and digest the things they are citing to.