josho

I'm sorry what⁉ what covid conspiracy theory has been proven correct?

I don't think it's superficial at all.

I genuinely find my apps harder to navigate now than I used to. Part of that is that I have far more apps installed today, but the uniform white borders also contribute. They make every icon look about 20% more similar, which adds just enough friction that scanning for the app I want takes a little longer.

Poorly executed icon shapes were distracting, but when they were done well they provided subtle visual cues that made the interface easier to navigate. I miss that more than I expected.

> It’s really disappointing to see the on-device models being limited to so few devices.

At first I thought it was the usual planned obsolescence. Then I realized it may be a true technical limitation. I suspect an embedding model is required to run on device in order to make several of the features work. Embedding models are small compared to LLMs, but, depending on their capabilities, could be the memory driver.

I loathe that I can't even rename the desktops.

Wouldn't it be great to have them named "Design", "Dev", "Productivity", "Games". Or whatever makes sense given your needs, instead of simply desktop #.

The meta data is useful.

Eg, When a prompt had a bad result and was edited, or had lots of back and forth to correct tool usage that information can be distilled and used to improve models.

And now imagine if you are focused on this for weeks you can likely come up with other ideas to leverage the metadata to improve model performance.

Similarly I worry about how these apps automatically update themselves. I know it can be done securely. I also doubt that these companies invest the engineering effort to do so.

The problem is that consumers are not savvy. They go to the store, and compare TVs based on features presented. Colors, refresh rate, size, etc.

Its only when they get home (and likely not even right away) that they discover their TV is spying on them and serving ads.

This is a perfect situation where government regulation is required. Ideally, something that protects our privacy. But, minimally something like a required 'nutrition label' on any product that sends our data off device.

This is the second app that I’ve tried with an AI voice for meditation. And frankly it’s off putting. The voice is great for other settings. But when my eyes are closed and I’m focused on nothing but the voice it stands out as negative.

Now I may tolerate that if you are significantly cheaper than the alternatives but that doesn’t seem to be the case.

Good analogy.

The solution, however, isn't prohibition or age restrictions; it's either regulating the algorithms or holding these companies responsible for the adverse outcomes their platforms contribute to. Safe harbor laws made sense when tech wasn't filtering/promoting content, now that they are influencing the material we see, these laws must no longer apply.

This may mean adopting a modern equivalent to libel laws. Something akin to: if an algorithm pushes false information, the company behind the algorithm can be sued for harm. Disallow terms of service that force arbitration or cap liability limits.

They didn't have a better definition of AGI to draw from. The old Turing test proved to not be a particularily good test. So lacking a definition money was used as a proxy. Which to me seems fair. Unless you've got a better definition of AGI that is solid enough to put in a high dollar value contract?

The EU should go even farther. Force hardware vendors to decouple from services.

Eg. AirPods work better with iPhones than Bluetooth. Why? Because of software integration. Apple Photos works better than third party photo management apps because of the OS to application integration.

The EU should require hardware makers to define compatibility tests and anyone that passes the compatibility test can become a drop in replacement for the vendor’s own apps.

This would increase consumer choice, competition, and reduce ecosystem lock-in. All of which will make things better for consumers.

4k jobs across the economy is far less than random variation in the stats.

Salesforce reduced their headcount in 2023 by 8-10%. Another reduction by 5% attributed solely to AI could be a half truth and the reality could simply be Salesforce driving an efficiency agenda.

Personally, I believe it will take a few more years for systems to be built. Once those systems are in place, then headcount reductions are going to come fast and wide. Or putting it simply think of it as exponential growth. Currently AI job displacements are small, but it's growing, and will continue accelerating in its growth.

Why? What problem does this solve?

With a car it’s common for people to not maintain correctly or to get in a major accident and not disclose.

What are the common factors that cause a computer to prematurely wear out? I can imagine there are lots of hypothetical risks, but how common are these? And how easy are they to mask?

The solution here shouldn't be technical; it should be legal.

If we rely on the technical path, Comcast can achieve the same by how many active IPv6 addresses are in use. Even if you aren't using your phone, the device is going to be constantly pinging services like email, and your ISP can use that to piece together how many people are at home.

If we rely on legal protection, then not only Comcast, but all ISPs will be prohibited from spying on their customers. Ideally the legislation would be more broad and stop other forms of commercial/government surveillance, but I can't imagine a world where Congress could actually achieve something that widely helpful for regular citizens.

Well if the demand isn't there then I suppose there's no need to have a 100% tariff on Chinese EVs, right?

The reality is the legislation was to push industry and it seems industries response was ‘nah dawg we’ll just lobby our way out of this one’