pslam

> The purpose it was originally intended to be used for doesn't make sense in a world where most new technology is just software, and software is incredibly easy to copy and duplicate.

The very first patent was to duplicate an existing process (the loom) and have a monopoly to produce it.

I keep hearing this argument from patent proponents, but patents have never in their history been ostensibly for good.

I am absolutely with you on this. I still think USB is an abomination, and the most modern incarnations only get worse — USB3 being essentially PCIe signaling but messed up because the USB committee got involved. They must ruin everything with their touch.

10/100 ethernet over a different cable and connector type would have worked out pretty nicely.

The specific case I was thinking of was the host suffering an incident where it is not possible or practical for its software to know where it left off.

For example, you get a kernel panic, or soft-reset for some reason. When you recover, you now have a bus in an unknown state, possibly mid-transaction, and if you pick the wrong order in which to bring the bus back to idle, you might wedge it or accidentally cause a side-effect (e.g overwrite a byte in an EEPROM).

Same. I really dislike I2C, but it's universal and it's been around for decades, and it's hard to avoid designs without it. I2C keeps causing these additional issues which the article doesn't touch on:

* No way to safely bring the bus back to idle from mid-transaction. By "safely" I mean not accidentally transmit an extra byte which could e.g overwrite EEPROM memory. There is no combination of transitioning the 2-wire bus from an arbitrary state back to idle which works in the general case. If it's important, you end up adding a dedicated reset wire.

* No safe, universal way to bring the bus from tristate, or low, to pulled-up. There are designs where this ends up being necessary. You end up with a spurious transaction, which may wedge the bus, or having to add a reset wire or buffer.

* The protocol is extremely hostile to devices with non-zero latency response. It's designed as a simple "Address this register and then immediately read out a byte in the next clock cycle". Works great for trivial devices, but for anything more complex it ends up needing a bank of register acting as a "proxy" to access the higher latency side of the chip. At this point I2C is an awesomely bad choice, but people keep doing this, because it's so universal.

The AI comes in as a way to make rich venture capitalists, sorry I mean philanthropists, part with their money.

If you cannot find a buyer who will take a cut of a $50MM valuation, then the company is in fact worthless. Liquidity is no barrier to this — there are plenty of financial instruments available to do this transaction, even if you can't actually transfer the asset itself. Taking on debt, for example.

Yes, it sucks, but forgive me if I don't shed a tear for someone who has to find a way to pay taxes on their $50M of assets. They absolutely do have options which aren't terrible.

If a "wealth tax" did pass, then expect countless startups who specialize in handling the arbitrage of this.

Uber's defense fails The Duck Test. They are describing a job, people doing a job, people paying for a job, and people taking a cut of the profits. They just don't use those words.

I suspect if/when this gets to a higher court, the whole thing will come crashing down, because to allow Uber's weaselly redefinition of common terms, would be to allow other classes of employment to similarly become unprotected.

I have no idea about this guy, but in my experience, people like this always seem to be pleasant, creative types on the outside. Every now and then in conversation they hold weird viewpoints you just can't reconcile with their projected image, and wonder if you've misjudged them.

They usually turn out to have an underlying wonky or absent moral compass, and abusive personality traits.

The outward image is a confidence trick, and that's how they got where they are.

> Between CarrierIQ and OTA updates/access, there is no such things as end-to-end encryption on a cell phone.

I don't think you understand what end-to-end means.

Yes, wearout rate definitely depends on temperature, but without reference to any actual data, this is what I mean by "no evidence".

Is it reducing 10 year lifespan to 9 years? 9.99 years? 5 years? Was it 50 year lifespan? It's pointless conjecture.

This article is actually describing how the Raspberry Pi 4 does NOT need a fan.

This is not the 1990s. It is perfectly acceptable and even advantageous to design for a high peak:normal load ratio, with thermal throttling. In this case, it allows for a compact, cheap, fanless design for the vast majority of users.

There is no evidence the heat dissipated will impact lifespan. It is common for the components picked out in particular (power supply, USB-C controllers) to be deliberately designed to run hot. They aren't made on the same process as the SoC.

I feel like there is a missing piece of the software/hardware design art here. There are many takes like this on the Raspberry Pi 4 design. Why only one ethernet? Why no fan? Why not more USB-C? Because it's $35 and because, perhaps, you aren't the target majority market. It's going to satisfy the vast majority of people, and those it doesn't have very simple and cheap ways to mod it so it does.

You are missing the entire point of the article if you continue to call these algorithms "AI". Inflating simple things like this to mean "AI" has led to the term being meaningless.

You are the example it is making.

> Backpropagation, which most researchers will agree is an AI algorithm, is a "simple algorithm".

Back-propogation is not an "AI" algorithm.

You are ironically doing exactly what this article is about.

This trick may break ordering and cache aliasing rules on many architectures. If you're dipping into this kind of thing, it needs per-architecture whitelisting.

I have routinely created firmware written in Rust which are just a handful of KB. The issues people are having are related to system-integration, such as static linkage, can be mitigated in cases anyone really cares about, and will disappear over time.