I challenge each and every one of you to make a pie by the end of the month.
I made one, for the first time in my life, last week. It brought me tremendous joy not only to make it, but to have something nice to share with friends.
My grandma made Platonically Ideal Pies, and I took up the art years ago. Mine, if I say so myself, are quite good, given that with Grandma's example I know what I'm shooting for.
I haven't made one for a few years, though - having a pie in my house is a recipe for me eating 5000 calories of pie and vanilla ice cream over the next few days.
When my grandma died a few years ago, I asked my aunts if I could have one of her pie pans. Apparently none of her other 17 grandkids thought to ask that - so I got all three (philistines!). Those basic metal pans are among my most cherished possessions.
In case you did not actually nail perfect, flaky crust the first time, that's a fun parameter to try to optimize. I finally got it at some point, and when I did, I realized that all those old cookbooks that said things like "use little water" and "keep the dough cold"---all the tricks where I thought "that has to be a myth"---turned out to be essential. The Joy of Cooking is full of old wisdom like this that has taken me ages to appreciate.
The "trick" to baking all kinds of things well for the first time is to follow the recipe fanatically. It says high protein flour? Use that, not all purpose flour. It says 500g caster sugar? Don't think that's going to be too sweet and add 400g, the sugar is where the texture comes from (and there's plenty of less sweet recipes to choose instead). It says make sure the dough is chilled? Chill the damn dough!
Once you've baked it perfectly to the exact recipe a few times, then you can start adapting.
Of course, there will come a point in your skill level where you will have the intuition to adapt recipes that you've never cooked before. But many people assume they can do that immediately, fail, then assume they can't cook and give up.
I will say though that the other biggest area where people fail is having a janky oven that can't maintain a stable temperature, or where you set it to 200C but it only reaches 160C. So an oven thermometer is a useful tool to buy.
I did this recently, and you know what I really loved about it? It's a great entry-level baking activity where the upside is that you have a pie (something you can gift or just eat!) and the downside is that you have a sort of cobbler.
You really can't !@#$ up a pie. Omelette is another good one. At worse you have scrambled eggs.
I mean, yes, at worse you burn your neighbourhood down and your dog runs away. But in terms of the more likely failure modes like screwing up the dough, breaking it, messing up how watery it is, etc. you can mostly just keep baking until it's done, mix it up, put into bowls, serve with ice cream, down the hatch.
I agree. When it comes to baking, making it tasty is mostly a matter of including the correct ingredients. Nailing the texture is the hard part, and that’s where technique and practice comes in.
I mean, it's hard to end up with something completely inedible but you absolutely can mess it up. Soggy wet pastry at the bottom is the biggest problem but there's plenty of advice around about how to avoid that.
Even broader, honestly. Make something culinary! It's amazing what the simple tactile experience of making something can bring when so much of our existence is doing things by proxy.
A fun hobby I picked up during Covid was trying to cook food from countries I had never been to - since traveling anywhere wasn't an option.
Pick a country, research what food it has that you've never tried, find a few online recipes and YouTube guides and give it a go.
This was a ton of fun. I have no idea if anything I cooked was even remotely like the authentic original, but it was still a very rewarding exercise.
If you live somewhere with a lot of international supermarkets (the SF Bay Area is great for those) it also gives you an excuse for a shopping adventure for ingredients.
I think you're under-estimating how much personal taste applies in that industry. Yes, there's a lot of free content but it's often low quality and/or difficult to find for a particular niche. The OF pages, and other paid sites, are curated collections of high quality stuff that can satisfy particular cravings repeatedly with minimal effort.
A big part of it also the feeling of "connection" with the creator via messages and what not, but that too can be replicated (arguably better) by AI. In fact, a lot of those messages are already being generated haha.
I was mostly hinting towards the 'connection' part of it, yes - I think that's really where the money is made more than anything else. That's the part that'll start killing the industry once some company tunes it in.
This is the dystopia of that pacified moon from "Mold of Yancy" by PKD but taken to the next level.
What's astonishing abut the present is that even PKD did not foresee the possibility of an artificial being not only being constructed from whole cloth but actually tailored to each individual.
Doesn't Grok allow users to create lewd content or did they roll that back?
Also, I suspect that we'll soon see the same pattern of open weights models following several months behind frontier in every modality not just text.
It's just too easy for other labs to produce synthetic training data from the frontier models and then mimic their behavior. They'll never be as good, but they will certainly be good enough.
Have you tried Qwen3 Coder Next? I've been testing it with OpenCode and it seems to work fairly well with the harness. It occasionally calls tools improperly but with Qwen's suggested temperature=1 it doesn't seem to get stuck. It also spends a reasonable amount of time trying to do work.
I had tried Nemotron 3 Nano with OpenCode and while it kinda worked its tool use was seriously lacking because it just leans on the shell tool for most things. For example, instead of using a tool to edit a file it would just use the shell tool and run sed on it.
That's the primary issue I've noticed with the agentic open weight models in my limited testing. They just seem hesitant to call tools that they don't recognize unless explicitly instructed to do so.
I think there are multiple ways these infinite loops can occur. It can be an inference engine bug because the engine doesn't recognize the specific format of tags/tokens the model generates to delineate the different types of tokens (thinking, tool calling, regular text). So the model might generate a "I'm done thinking" indicator but the engine ignores it and just keeps generating more "thinking" tokens.
It can also be a bug in the model weights because the model is just failing to generate the appropriate "I'm done thinking" indicator.
I was testing the 4-bit Qwen3 Coder Next on my 395+ board last night. IIRC it was maintaining around 30 tokens a second even with a large context window.
I haven't tried Minimax M2.5 yet. How do its capabilities compare to Qwen3 Coder Next in your testing?
I'm working on getting a good agentic coding workflow going with OpenCode and I had some issues with the Qwen model getting stuck in a tool calling loop.
Minimax passed this test, which even some SOTA models don't pass. But I haven't tried any agentic coding yet.
I wasn't able to allocate the full context length for Minimax with my current setup, I'm going to try quantizing the KV cache to see if I can fit the full context length into the RAM I've allocated to the GPU. Even at a 3 bit quant MiniMax is pretty heavy. Need to find a big enough context window, otherwise it'll be less useful for agentic coding. With Qwen3 Coder Next, I can use the full context window.
Yeah, I've also seen the occasional tool call looping in Qwen3 Coder Next, that seems to be an easy failure mode for that model to hit.
OK, with MiniMax M2.5 UD-Q3_K_XL (101 GiB), I can't really seem to fit the full context in even at smaller quants. Going up much above 64k tokens, I start to get OOM errors when running Firefox and Zed alongside the model, or just failure to allocate the buffers, even going down to 4 bit KV cache quants (oddly, 8 bit worked better than 4 or 5 bit, but I still ran into OOM errors).
I might be able to squeeze a bit more out if I were running fully headless with my development on another machine, but I'm running everything on a single laptop.
So looks like for my setup, 64k context with an 8 bit quant is about as good as I can do, and I need to drop down to a smaller model like Qwen3 Coder Next or GPT-OSS 120B if I want to be able to use longer contexts.
After some more testing, yikes, MiniMax M2.5 can get painfully slow on this setup.
Haven't tried different things like switching between Vulkan and ROCm yet.
But anyhow, that 17 tokens per second was on almost empty context. By the time I got to 30k tokens context or so, it was down in the 5-10 tokens per second, and even occasionally all the way down to 2 tokens per second.
Oh, and it looks like I'm filling up the KV cache sometimes, which is causing it to have to drop the cache and start over fresh. Yikes, that is why it's getting so slow.
Qwen3 Coder Next is much faster. MiniMax's thinking/planning seems stronger, but Qwen3 Coder Next is pretty good at just cranking through a bunch of tool calls and poking around through code and docs and just doing stuff. Also MiniMax seems to have gotten confused by a few things browsing around the project that I'm in that Qwen3 Coder Next picked up on, so it's not like it's universally stronger.
Thanks for the additional info. I suspected that MiniMax M2.5 might be a bit too much for this board. 230B-A10B is just a lot to ask of the 395+ even with aggressive quantization. Particularly when you consider that the model is going to spend a lot of tokens thinking and that will eat into the comparatively smaller context window.
I switched from the Unsloth 4-bit quant of Qwen3 Coder Next to the official 4-bit quant from Qwen. Using their recommended settings I had it running with OpenCode last night and it seemed to be doing quite well. No infinite loops. Given its speed, large context window, and willingness to experiment like you mentioned I think it might actually be the best option for agentic coding on the 395+ for now.
I am curious about https://huggingface.co/stepfun-ai/Step-3.5-Flash given that it does parallel token generation. It might be fast enough despite being similar in size to M2.5. However, it seems there are still some issues that llama.cpp and stepfun need to work out before it's ready for everyday use.
Isn't it just the usual feedback loop that happens with popular podcasters? They have connections and get a few highly popular guests on. As long as their demeanor is agreeable and they keep the conversation interesting other high profile guests will agree to be on and thus they've created a successful show.
I guess it was just a poetic riff on Tinder for AI agents. It seems like one of the more profound questions around AI and the singularity. One AI gaining sentience would be a big deal, for sure, but two self-aware AIs that could produce an offspring — that would be quite something.
The knee-jerk reaction reaction to Moltbook is almost certainly "what a waste of compute" or "a security disaster waiting to happen". Both of those thoughts have merit and are worth considering, but we must acknowledge that something deeply fascinating is happening here. These agents are showing the early signs of swarm intelligence. They're communicating, learning, and building systems and tools together. To me, that's mind blowing and not at all something I would have expected to happen this year.
> These agents are showing the early signs of swarm intelligence.
Ehhh... it's not that impressive is it? I think it's worth remembering that you can get extremely complex behaviour out of conways game of life [0] which is as much of a swarm as this is, just with an unfathomably huge difference in the number of states any one part can be in. Any random smattering of cells in GoL is going to create a few gliders despite that difference in complexity.
I made one, for the first time in my life, last week. It brought me tremendous joy not only to make it, but to have something nice to share with friends.
reply