HN user

TGower

130 karma
Posts0
Comments73
View on HN
No posts found.

I can't see where one can opt-out of this new behavior and into the existing behavior, only a description of the new behavior's bypass (which is not the same thing at all)

I don't understand this, the ability to bypass new behavior in settings menus is basically the defenition of a new feature having an opt-out. Can you elaborate?

This is a wild misrepresentation of the situation. Saying there is no opt-out is just false, they even provide the information on how users can opt-out. The "mandatory 24 hour cooling-off period" is also misleading, it's easy to bypass the cooling-off period with ADB.

Gemma 4 on iPhone 4 months ago

These new models are very impressive. There should be a massive speedup coming as well, AI Edge Gallery is running on GPU, but NPUs in recent high end processors should be much faster. A16 chip for example (Macbook Neo and iphone 16 series) has 35 TOPS of Neural Engine vs 7 TFLOPS gpu. Similar story for Qualcomm.

Prepaid mobile carriers offer "free" phones with one month of service. $25 a piece for Samsung a16 4-pack from Metro, if you only want 1 you can do $45 for a moto g power (8gb RAM) from Straight Talk. I've been using these as alternatives for SBC projects, cheap, battery and screen can be nice, most GPIO needs can be achieved by slapping on a $2 ESP32 clone over usb-serial. The supply will probably dry up soon, I think they might be eating a loss on just the RAM BOM.

The easiest way would be to setup some cameras and start recording everything. Gemini could already sort that into events you could query. If you have privacy concerns, at the current pace of progress, local LLMs should be up to the task soon if they aren't already.

I still don't get how they are driving away devs. It's super easy for us to click the setting. If I urgently need to test my app during the 24 hour waiting period, I can just adb it on my device.

Based on the reaction here, it's obvious I'm missing something here, but I just don't see any real reason devs are feeling like they are being driven away. It's hardly more of an inconvenience than enabling developer mode, and I feel like we all get why they hide the developer settings menu behind that.

Really, there are apps that will intercept and exfiltrate your bank one time code sms that are just sitting on the play store? First I'm hearing of this, what's the name of one?

It really seems like they are doing a lot to appease the tiny minority of us power users, adb load unaffected, one time toggle in settings to opt out, no change to alternative app stores as long as the apk was built by a verified developer. Crazy how harsh the sentiment is here, there are real people being harmed by scam apps intercepting sms one time codes and this will reduce the rate of that happening. It's not like we can't sideload anymore, though a lot of comments here seem to be implying otherwise.

Nanite is a good counterexample, very impressive and innovative technology. Even more impressive that they released the technical details instead of going the software patent route. I think trying to leverage the war chest to go after Steam's monopoly was exactly the type of adventurous plan you are talking about, the safe play would have been to make minimal investments and continue churning out games hoping for another big hit.

With those past examples the majority of people thought "It's not my problem, I won't be affected", and they were mostly right. With an automation wave of the scale needed to get to effectively no jobs, that's just not the case. They will see that it is indeed coming for them, their friends, and family. They will act accordingly. Altruism not required, just self interest.

a planet that is breathable and isn't burning

This is a perfect example of something that can benefit greatly from abundant goods and services. Driving the cost of solar panel manufacturing, supply chain included, and deployment. Enabling continuous monitoring and fast response to GHG leaks or forest fire starting. Reforesting efforts. There are so many ways in which the application of intelligence and labor can help us here, and AI can vastly grow the supply of intelligence and labor.

More free time?

Yes! Time we can reclaim from the mundane chores of life to do with as we choose! How could you not want that?

Right now it is difficult for the average person to put themselves in the shoes of a homeless person. There are a litany of ready made excuses not to do so: "Oh it's the drugs", "They aren't even going to the shelter", "They must have mental illness", a variety of ways to say "I could never end up like that, if it was me I'd do better and pull myself out". These excuses evaporate in the face of a real automation wave where a large portion of friends and family you know to be hard working and intelligent are finding it impossible to find a job.

This is true. It's by no means garunteed that we will get to a point where effectively all the jobs are being automated. If we eventually get there, it seems likely the path will be gradual and prosperous enough that we can handle the transition in a way that provides for everyone. The dangers of the alternative route are real, but hopefully obvious enough that we can collectively avoid them.

How does AI truly benefit the average joe?

Automated production of goods and services means more goods and services to go around. From cheaper prices on all of the things people already buy to unlocking new classes of products like actually useful robotic helpers. Increased pace of development and reduced cost will make many niche products economically viable, essentially the maker movement on steroids.

This seems like a good solution that will put a sizeable dent in scam success rates while not actually removing options for developers and power users. The added friction will make some people bounce off F-Droid and the likes which is unfortunate, but the wins here in scam prevention are much bigger than the losses in onboarding power users.

It wouldn't make sense to have the LLM try to do the target recognition, trajectory planning, or motor control. It might make sense to have the LLM at a higher level handling monitoring of systems and coordination with other instances, to provide more flexibility to react to novel situations than rules bases systems.

With a chess engine, you could ask any practitioner in the 90's what it would take to achieve "Stage 4" and they could estimate it quite accurately as a function of FLOPs and memory bandwidth. It's worth keeping in mind just how little we understand about LLM capability scaling. Ask 10 different AI researchers when we will get to Stage 4 for something like programming and you'll get wild guesses or an honest "we don't know".

Try out running different edge detection algorithms on that image and you will see that there is a lot of disagreement amongst them in the feathering region. Exploring what the differences are, and how the algorithms lead to those differences helps build intuition about the range of things we might call an "edge", and which algorithm is appropriate for a particular task at hand.

"elected for no obvious reason" isn't quite right, as a test image for computer graphics it has regions of very high frequency detail and regions of very low frequency detail which make it easier to spot various compression artifacts, and it makes a good study for edge detection, with both very clear edges along the outline, but more subjective edges in the feathering.