pzo

in many cases often its faster just to switch from debug to release - compilers are good to vectorise many loops. Worth to give it a try before rewriting clean loop/code into SIMD/NEON.

You can as well make it simpler and use those skills:

https://github.com/software-mansion/argent

or

https://github.com/callstack/agent-device

both callstack and swmansion are mostly react native shops but those should work even in native ios/android as well

I think they would still fail with MeeGo, same like Blackberry failed even though they had really good OS. I remember was going to their conference in amsterdam and were giving their new Blackberry Tablet for free to attract developers to make any app.

If Nokia and Blackberry team up they maybe would succeed. Nokia Mobile still could be around if instead choosing Windows Phone would release Android phone back then.

I stopped reading after seeing they compared only with Whisper Small, Base, Tiny

This is useless test and benchmark when you have these day Whisper-V3-Large and Whisper V3-Turbo that you can faster than realtime on 5 years old macbook on apple sillicon (ANE). They didn't even compared to parakeet v2 or parakeet v3. And only english language...

but also: "ADDW systems use cameras and sensors to track a driver’s head position, eye movements, and gaze direction"

and

"only recording and retaining data on the device that is necessary for the system to function"

Not sure why such stuff would need to be retained to function. Also if they need to track head position and gaze direction those definitely have to be higher resolution camera and/or pointing directly at your head - so at least capable to store biometric information and hacked remotely since more cars are having telemetry.

why not instead fork repo just in case but still use better auth until proven wrong? In case they go evil you just build from you forked one. At that point you would still have to either maintain your own auth or better auth fork.

With current AI your agents probably still will be better with maintaining a fork. Auth libs have pretty limited API surfaces comparing to e.g. ui frameworks.

Even funnier that CUDA in polish means “miracles”

FWIW there are today many more alternatives with better license. Here is a good meta repo for object detection with different model variants:

https://github.com/LibreYOLO/libreyolo

try to ask even grok about some stuff happenning right now in middle east or related to epstein files - its more and more censored and only sometimes will answer if you ask know what detailed question to ask. One year ago grok wasn't that bad and its supposed to be the less censored.

You also need to have time to play all those great games, watch all those great shows and read all those books.

There is a lot of software that are constrained by human eyeballs (entertainment or ads monetized)

Firstly, we might finally become productive enough to exhaust the world’s appetite for software. I don’t see any evidence of this happening, but if somebody wants to make this argument, they should be clear about why this time is different to the entire history of the computer industry so far.

Not all software but how many movies you can watch, books read? Similarly how many games you can play? How many ads you can watch during a day?

Eventually projects will have to be profitable. And even if everybody will make a great game most of them won’t be profitable because you competing for limited eyeballs time

Quite a good release although not sure why they invest so much time into their ONNX engine. I don't think they have enough stuff and big pockets to compete with ONNXRuntime, CoreAI, ExecuTorch, LiteRT.

I'm happy they added option for ONNXRuntime. I wish their cv.dnn was mostly that unified wrapper around many different backends (ONNXRuntime, Executorch, LiteRT, CoreAI) and maybe just some tooling around it (performance metrics tools, model downloads etc). Transformers(.js) approach looks better for me.

Wish they also invested more time into better production ready Camera I/O (for mobiles, device/format discovery, manual settings, depthmap support, etc) and better Highgui that could use different backends (skia, webgpu) and on mobiles.

how are supposed to use TensorRT on iOS, iPadOS, Android or even Web? Production is not only cloud.

seems they planning to replace it but overall now I'm really confused about this and mlx and coremltools. They should do better work explaining the benefits (and cons) of it and any feature parity between coreai, coreml and mlx.

TLDR; this is just Claude Haiku altrenative, you can probably skip whole article.

I did have few days ago conversation with AI to research how the economical environment was just before Great Depression and this doesn't look good if true. Definitely feels like someone try to dump their bags to retails investors and main street.

- Trailing P/E: 1929 peak was 32.6 vs. 32.67 today.

- Shiller PE (CAPE): 1929 peak was ~30 vs. 42.66 today (2nd highest in history).

- Buffett Indicator: Market cap was 124% of GNP in 1929 vs. 259.6% of GDP today.

- Margin Debt to GDP: 3.0% in 1929 vs. 4.1% ($1.304T) today.

- Systemic Risk: 1929 margin debt was 10-12% of total market cap vs. 1-2% today. Modern leverage has structurally shifted from retail to sovereign debt and shadow banking.

how many packages have you shipped so far to space? SpaceX could disappear tomorrow and most people wouldn't notice. Your satellite TV might get slightly more expensive. Those rare people that don't have LTE internet access and need starlink are exception.

I had the same question when reading website - really poorly explained. Not sure why they also advertise that those genetically modified males are non biting and harmless - all males mosquitos are like that.

If you are a heavy user of coding agents these plans are a fantastic deal. I just ran the ccusage tool on my laptop to get an estimate of how much I would have spent if I were to pay for API tokens in the past 30 days and got

You think this is fantastic deal only because they use similar like tricks where they inflate the price and tell you something supposed to cost $1000 but they have this today promo for $100.

I was there too and paying for a while. Few weeks ago I tried DeepSeek V4 Pro - expected its gonna be shit but its actually pretty good.

The deal is I pay daily ~$1 for DSV4-pro for ~100M API token usage. And they probably not getting broke because >90% of those token in practice is cache read and they very well optimized for that.

you doing probably something wrong, I used Deepseek v4 pro with opencode and in a day used 100M tokens for ~$2. Majority of tokens are cache tokens and those are extremely cheap in deepseek bordering free.

I compare ultra wealthy to blackholes - overtime they accumulate more and more mass and reduce mass elsewhere. But even in nature we have Hawking radiation (which leads to black hole evaporation). So for me wealth taxation is similar like this slow black hole evaporation which seems fair.

at least all financial assets (stocks, etc) are easy to assess value so why not start with that? same with gold, silver etc. Some minimal amount you can make it nontaxable to reduce administrative burden.

they also wiped out my whole plugin, themes, config setup...

There is a reason in capitalism we have anti-monopoly law or preventing dumping prices because those often leads to monopoly. So yes for sure you can kill your competition by just dumping money and loosing profits.

both USA and EU are stagnant but you have to go to East or SE Asia to see what it means. Salaries and GDP is not always the best measure of growth.

I think you are collapsing too many different causes into a single explanation.

Yes, consumerism probably influences expectations and lifestyles. But many of the things I mentioned are not just superficial manifestations of consumerism - they are structural economic and social changes.

When people say they "cannot afford" more children, they usually do not mean literal starvation or inability to keep a child alive. They mean they cannot afford the living standard that modern society effectively requires or expects for a family with multiple children.

I mentioned, in Poland when I was growing up, it was normal for 3 kids to share a small apartment and for grandparents to help raise children. Today, many young adults had to move to larger cities for education and work, far away from their families. That removes a major support system.

Now both parents usually need to work, which creates additional costs: larger housing near jobs, childcare, kindergarten, transportation, often even a second car. These are not just luxury consumerist indulgences but practical requirements of modern urban life.

Materially, we're the wealthiest we've ever been in human history.

But wealth being higher on average does not mean family formation became easier for the middle class. Housing costs in major cities relative to income are a huge factor, especially for people who are not poor enough to qualify for assistance and not rich enough to comfortably absorb the costs.

it's nondeterministic because we chosen it by having higher 'temperature' in settings. I bet if you run open weights model with temperature 0 and on the same device the same prompt and turn off parallelism you will have more deterministic result (excluding some floating point operations).

This is very simplistic and I would say there is more reason than only consumerism. People still might have kids they just have it less - they are happy to have only one kid because they fill fulfilled and also they cannot afford 2 or 3.

Standard and expectation also increased and even thought I grew up with 2 siblings in 2 bedroom apartment in Poland today nobody would want that - or good luck finding a partner that want that. You would expect to have house or at least 3-4 bedroom apartment to raise 3 kids.

Today also probably you need 2 cars instead of 1 family car because your partner also have to work. You probably also need extra money for babysitter or kinder garden because again your partner is working and probably less likely your parents nearby to help since most young people had to move to big cities to get a job.

Sorry, I'm a bit tired of acting like Germany only got the history of being on the west side of the iron curtain. It got both treatments.

Well glass that has half of water is still better that glass fully empty. Poland didn't get any war reparations and after being more destroyed during the war than germany (warsaw burned to the ground) and pretty much occupied for many decades after the war then how polish companies supposed to compete with any western economy including germany?

Many (especially Germans) try to wrap it falsely as charity and throwing numbers without context.

It is not true for the entire EU budget: the 2014–2020 MFF allowed up to €959.99bn in commitments, so Poland’s €77.6bn cohesion allocation was about 8.1% of the whole EU long-term budget.

For context this is like 0.7% of yearly German overall public expenditure go to Poland. And this money per year is also like 5% of state budget spending.

a big share of this money goes back to foreign companies in form of sales and contracts.