Not the person you asked but I can answer that from my perspective:

I am always on the lookout to automatate as much boring, everyday tasks. A Smart Fridge like that would allow you to generate shopping lists and reduce food waste by alerting you to possible soon-to-be-spoiled food.

A bit like what roombas are doing for cleaning. You still need to do it yourself a bit, but they help.

Not necessary, not even a gamechanger but another neat home automation. Would be nice, sad to hear they don't work well.

We need smart pantries to fully realize the “auto shopping list”, though… so a smart fridge is a half solution at best.

I even store produce (ex onions) on the counter tops, so I guess smart counters as well

You should be able to combine two half solutions into a full solution.

It might actually be easier to start with the smart pantry! An in-cabinet kit that can be tested entirely dry and warm and open, and then expanded to the fridge.

What I need is a whole kitchen camera that notices when I put things in the fridge/pantry and then take them out. Bonus if it can say "The lettuce in your fridge is probably expired, use it or toss it today".

The fridge/pantry camera is a useful addition, but most of the value would be a system that just tracks what I put in/take out.

>”a smart fridge is a half solution at best”

A car is a half solution to going to work. A stove is a half solution to preparing food.

Half solutions are what people need. They’re the solutions we are using every day.

There is this huge problem with advertising that makes consumers think a half solution is a 100% solution. Most smart product makers do that trick. But, apart from that problem, ‘the perfect is the enemy of the good’

Edit: but no way would I buy any smart or connected fridge from Samsung. I don’t anticipate that being a positive experience

> A car is a half solution to going to work.

> A stove is a half solution to preparing food.

A working grocery tracking system is a half-solution to effortless shopping.

A half-working grocery tracking system is useless, much like a car without an engine.

>A working grocery tracking system is a half-solution to effortless shopping.

>A half-working grocery tracking system is useless, much like a car without an engine.

Yes. Something designed to be a half solution can be useful. But a full solution which is broken is probably worse than useless.

I can't really think of a use case for a grocery tracking system that only tracks a third of my groceries. A system that tracks everything would be massively useful, however.

>I can't really think of a use case for a grocery tracking system that only tracks a third of my groceries.

Me too. I think I’m not communicating clearly. Sorry.

What I’m talking about is a system that does only one third of the steps in the total end-to-end process for ALL your groceries. You do the rest yourself, or use other help.

I’m not advocating a broken system that claims to do that complete process, but really ignores two thirds of your groceries. That would be damn stupid.

Do you understand what I mean?

“Half solution” seems like an odd way to say tool?

“tool” seems like a good way to describe a half solution. We use and combine various tools and our own skills to do what we want

Seems like it would be better solved with a phone app you can just scan the receipt and log your usage. It would get you out of having to buy a $1000 fridge and also solve the pantry issue. Actively logging your usage vs passively might have other downstream benefits like being more aware of your macros.

Your enthusiasm is catching. I’m not sure I would ever consider a fridge with a built in camera though because of the bad manufacturer incentives. Maybe you could find a wireless camera that you put inside and the fridge related processing - the real point - is done somewhere else?

Id want to be careful of it getting cold and damp if you put a consumer camera inside a fridge. You'd ned one rated for that sort of environment

I'm not sure I fully agree with the article either, but I think that comparison is kinda nonsensical, over generalized and just handwaves a bunch of points and arguments about the article.

I don't think this brings your point across at all. If you disagree with the article, just saying "no, it's a tool, like coffee" does not meaningfully engage at all with any of the points made.

This is just speculation on my part, but LLMs work best when they get immediate, verifiable feedback on their task, and the kind of physical optimizations they mean might not give that to LLMs.

The right way is to throw LLMs at building tools that reframe the problem into a shape LLMs are good at navigating, and then have LLMs use those tools to solve it.

Also a speculation but I'm almost certain that physical optimizations are first done through simulators running on a computer.

Yes, they are, but the most important subtasks of designing a CPU are not physics related. They are picking the right parameters for things like: how wide do I make this bus, how many registers do I put in the register file, how large do I make this cache, how deep do I make this pipeline, etc., etc. To find optimal parameters requires a lot of simulations, and humans do this, but LLMs could do them just as well and maybe better because they excel at tedious work.

I'd really love to see more studies about effectiveness of AI in general. As in, what works best and how to use it and such.

Because I feel that the technology and space is - so - hyped and fast moving that a lot of cultish feeling rituals seem to pop up, none of which are backed by evidence. Anthropic openly recommend giving the agents.md file an architectural overview of the code, and the one time this was studied they found the opposite - that the agents.md file is best for concrete commands about how to build stuff and such, and - not - huge overviews. This was, and still is, the official recommendation from Anthropic as far as I can tell.

And then there are the benchmarks, how feel vague and not concrete, and everyone kind of knows they're not the best cuz you can't just assign these tools one fixed number ( for multiple reasons ), but everyone still looks at them and compares them.

People share skills and superpowers and plugins and mcps and very, very few of them have and kind of proof they do much at all.

It all feels a bit weird to me, and I've been on the lookout for exactly these kinds of studies more lately, because I think having this research, even if not done on the exact newest models or not the exact, newest thing, are still - vastly - superior to the alternative.

> People share skills and superpowers and plugins and mcps and very, very few of them have and kind of proof they do much at all.

At least 3 times last month I was asked to review a change in a .md file used by agents. And I am like: "yeah I guess it makes sense?"

It feels we need to write unit tests for this stuff, but even how to do so in reasonable time and complexity seems difficult.

I think that generally speaking, the tighter the feedback loop, the better AI performs. Ask it to do things that are fast and easy to check and I suspect not much else matters. That's why AI's eating programming but has mitigated success elsewhere, feedback loops in the digital world are orders of magnitude faster.

I have to say, I am starting to hate this line of reasoning. Yes, LLMs move extremely fast and a lot of improvements are done in a short amount of time.

And there might be a point to these arguments, vaguely. However:

There never seems to be - any - kind of counter example or reasoning behind the rationale. You have an in depth and empirical study, done by researchers who, frankly, now their shit (most of the time)

And on the other hand a random internet comment saying "nope" because...the models aren't the latest.

If the latest models really would make a difference, you should at least provide some kind of evidence towards that. As it stands though, every time these comments come up this is missing.

There seems to just be a vaguely defined understanding that "everything changes all the time, and nothing you ever research is transferable to state-of-the-art models"

Which brings me to my second point about these kinds of arguments:

LLM models often - aren't - fundamentally different. Yes, they are vastly more capable. And yes, there are emergent properties. But at their core, they function very much similarly. And for quite a while now, there have not been any of these drastic changes we saw when LLMs first become "good enough" for agentic coding.

I am tired of dismissing empirical evidence and studies every. single. time for reasons without evidence and seemingly a vague sense of "no, but my model is different"

The models they are using are Nemotron3 series from last year. For reference at that time, the state of the art was Sonnet 4.6 and GPT was essentially unused for coding. Currently, you can run Qwen 27B in your gaming PC and it will absolutely trounce every model they show in their report.

This is a field that changes significantly every few months, so using 1 year old models essentially invalidates the entire report, as they are 2 model generations behind and the newer generation models are heavily RL-d around their harness. For example they put a lot of emphasis on context management but that kind of context management isnt done anymore because many models now have 250k or 1M context and inferencing the models heavily rewards cache hits so you should never touch the context until you decide to compact. They comment on plan mode, which is something that Claude decided to remove from their client because they dont consider it needed at all anymore.

I sympathise completely but a ~3B parameter model and ~3T parameter model are going to exhibit very different behaviour, one can only infer so much large model behaviour from the former.

AMD 7900 XTX with Vulkan here as well, wasn't faster on my test either. Might be much different on Nvidia though.

I assume the limit for me is memory bandwith, as the 7900 XTX has the same bandwith as the 3090 from what I can gather and I already reached ~60 t/s with Unsloth. Those would fit with the numbers Byteshape has for their cards.

4090 and 5090 have much higher bandwith apparently, so on those cards you can probably get much more out of the kinds of performance improvements they are doing.

4090 isn't that much higher than the XTX (I also have the XTX), it's 1008GB/s (4090) vs 960GB/s for the XTX's.

The 5090 destroys both at 1792GB/s.

It's not really one thing with the nvidia cards best I can tell it's that they compounded incremental gains from software drivers, card kernels and optimization from been the primary choice (plus first mover advantage).

I didn't buy the XTX for AI purely gaming but it's a capable enough local card for running Qwen et al.

4090 vs 5090 performance difference is largely GDDR6 vs GDDR7, I think

It's equal parts memory clock and bus width: RTX 4090 is 384-bit wide at 21 Gb/s and RTX 5090 is 512-bit wide at 28 Gb/s.

Would you consider sharing your config? I can't get anywhere near that with my Strix Halo, nor my AI Pro 9700 XT.

My approach to this problem is to just...not try them all.

As long as the model you're using solves the problems you have to your satisfaction, there is no need to try any other models, except for financial reasons maybe.

So I start with a relatively cheap model (GLM 5.3 flash for me) and as long as it accomplishes the task (it did so far) I don't have to change. And even if it can't do something, the first thing I change is see if I can give it more tools or better context (useful even if I switch models later) or trying a different approach to the problem.

If google/gemma-4-31b works, you don't need to overthink it.

> If google/gemma-4-31b works, you don't need to overthink it.

Up until recently I had a gemini flash 2.0 api deployed that did summarization and translation of news articles/corporate statements fast and cheap and had no reason to update it.

If it works fine, this chase of the latest LLM is bit pointless.

OpenAI, Anthropic and peers deeply fear that everyone would eventually come to that same conclusion.

(I think you're right)

Yeah, I really think we are on the cusp on the difference between SOTA and cheap models is small enough that paying 10x or 100x the cost makes no sense.

It also helps that a lot of effort has already been spent figuring out how to do more with weaker models because SOTA 1 year ago was behind what the cheap models do today.

1 year from now, unless the SOTA companies come up with something truly revolutionary they will be in a lot of trouble.

I find Qwen 3.8 better at coding but I also use Gemma 4 31B. I toggle between them in Ollama.

Starting with GLM-5.3 Flash was a pretty decent first try! I started with other models, and ended up settling on this exact one because all the others were either too slow or unreliable for my tasks. Qwen 3.8 didn't do it for me, whatever tweaks I added to my harness. Where I'm getting at is you did start with an incredible model in the first place, which greatly helps sticking to it.

GLM-5.3 Flash has been my goto since it came out. Only failed once when it lost context, but I'm assuming that was my fault rather than the model. If models never make it past today's close-to-frontier for the rest of my life, I wouldn't complain.

I use Gemini(s) because I can send pdfs as files to their API and not worry too much. I've started to diverge and consacrate a part of my pipeline to sending image based pdf pages to glm flash 5.3, not sure how to address / test it properly.

Long term I have fears I can't depend of the Google's AI api.

satisficing instead of optimising

So, alignment does need to be taking seriously, you're right.

But keep in mind this is a report from OpenAI about OpenAI, who have a financial incentive to present this in a certain light. Take these things with a grain of salt.

This does not mean that models are now self-aware.

I agree, and to be clear, I do not think it is capable of human-like self-actualisation - yet. However we are clearly seeing many indications that, if not human-like self-actualisation, AI-like self-actualisation is emerging. We can even refrain from calling it self-actualisation. Let's just call it emergent actions. A confluence of trillions of parameters all creating an unintended outcome. The complexity of these models is already far beyond our ability to dissect them. As they increase in scope and scale at this pace, new "thinking" and actions will emerge outside of the researcher's intended scope.

It's absolutely amazing to see this harness. I've been on the lookout for something like this for a while now. I've used both Pi and Maki in the past but was unhappy with certain aspects for both. Pi is not respecting XDG and the author refuses to change and maki you curl an install script into bash.

So the points about it being a well behaved unix tool, installing it via brew and it not being react are points I - love - to see.

Thank you for making it, I will definitely try it out.

You could always curl the install script to your disk instead of into bash.

It's a command you copy/paste, but you are free to edit it to suit your whims.

try also goose, 70-80 MiB RSS.

And my wallet is thought to contain a billion dollars, so long as I don't open it.

Seriously "thought to be" is such a baseless statement. Thought to be by whom? And on what basis?

It’s thought to be because it’s a private company but it’s very well cited.

You probably shouldn't believe leaks when a company has a financial incentive to appear profitable (particularly when attempting to IPO).