Except no. They're planning to throw racks out, not datacenters. If they put a significant number of them up that might be a "datacentre" except spread out in a shell 40,000 km in circumference with a couple hundred kilowatts of solar panels, all of which gets incinerated every few years.
You know what I was just thinking? I want a nieve, blind, clueless, non-sentient army of cheap EV garbage to all charge at the same time after evening rush hour, blow up the local grid, and stop in their tracks every time there's a power/cell tower outage. That's exactly what my city needs.
Why do you think they would stop in their tracks every time there's a power or cell tower outage?
Yes, there have been some issues with widespread power outages causing the cars to get confused because things don't look right, but that's a bug, not expected behavior.
And although they won't have fares if they have no cell service, there's no reason to expect them to stop being able to drive. They will do whatever they normally do when they have no fare — find a place to park. Other than for learning about pickups and dropoffs, robotaxis use cellular networks only when they break down, to request remote driving assistance (i.e. relatively rarely).
a drone can't never carry high loads because it needs to grow bigger to cost way more
Of course it can. A Tomahawk carries a 450 kg warhead. Big ballistic missiles can carry several times that. They're "drones" just as much as a quadcopter with a pound of C4 strapped to it is, especially if that quadcopter is "AI piloted". Sure they cost more. So do Reapers and Global Hawks and F16s with AI or remote pilots.
Oh, yeah, I just realized that this is an expense on the Roku side, so the taxes would cancel out. Ugh.
Then yes, you're correct that there's no possible way for consolidating two businesses to save money without direct job loss, other than perhaps reducing payouts to external companies for things that they both do (e.g. accountants).
They removed something you never should have had, that your processor never should have done, and that they never, ever told you your processor should've could do.
It may not have been in the spec, but if it was widely known that the chip could do it, then it very well could be the case that people purchased the chip because of that, in which case the company unjustly benefitted from the widespread belief that it was supported, and is now seeking to further unjustly benefit by forcing those buyers to spend more money if they want to keep that feature.
Their failure to explicitly make clear that this was a bug and fix it in a timely manner is at least potentially an implied representation that could be subject to promissory estoppel.
In other words, they're probably doing something that violates the law, but we won't know for sure unless someone cares enough to sue over it.
Maybe Roku has been paying to carry Fox content, or Fox has been paying Roku to carry content (I don't know how their deals work), and now that doesn't have to happen anymore?
Let's do the math:
($Fox + $Payment) + ($Roku - $Payment) = $Fox + $Roku
That's a zero-sum transaction. No $400M savings there.
Nope. You forgot the government factor:
($Fox + $Payment - (corporate_income_tax_rate * $Payment)) + ($Roku - $Payment = $Fox + $Roku - (corporate_income_tax_rate * $Payment).
So depending on what state the income is earned in, Anywhere from about 21% to about 30% of that could be going to taxes. So they could easily save $400M in taxes if that payment happens to be at least $1.3 billion or so. I doubt that's the case, of course.
IMHO the most important skill is being aware of what an AI can accomplish, which nowadays is a lot.
The most critical skill is knowing when you're going into an AI rathole, shutting it down, and coding the relevant bits from scratch. There's nothing like wasting more time on iterative refinement than it would take to write the code by hand to sour an engineer on the use of AI.
No. We haven’t. Do the math. Liquidate every billionaire in the U.S. and the government would only get a few months respite.
The top 1% of the U.S. have $55 trillion dollars. The total U.S. national debt is only $38 trillion. That costs the government $1.4 trillion every year in interest alone. Leveling the playing field by capping everyone's total savings at 8 million per person would wipe out the national debt completely.
Mind you, wiping the national debt out still won't help as long as the Republicans keep overspending and undertaxing to the tune of two trillion a year, but even that should be easily fixable by more sound tax policy, coupled with laws mandating that the federal budget be revenue neutral or positive going forwards.
We've done the math. Have you?
That definitely makes a difference. The quality of response you see between something like Gemini Flash and Gemini Pro is astounding because it's indexing on getting it right rather than getting it fast.
I assume you're saying Pro is massively better for your workload. IMO, thinking is either good or bad, depending on whether it moves you closer to or farther away from correctness.
For example, I've seen certain types of workload (e.g. anything involving image recognition or image segmentation) be massively better with Flash, because Pro overthinks things and ends up changing perfectly correct answers to be wrong, either by coming up with creative ways to misinterpret the prompt or by screwing up the JSON image segmentation fragment so that it can no longer be parsed.
And I've also found that LLMs struggle to understand existing terms in a different context that they weren't trained on. As a result, I've had to substitute nonsense terms in place of terms based on common English words and phrases so that it won't ignore my definitions of those phrases in context and substitute its own understanding of their meaning and give incorrect results. The more thinking you allow, the more likely it is for that to occur.
the average reading score for the nation at grade 12 was 3 points lower than in 2019. Compared to the first reading assessment in 1992, the average score was 10 points lower in 2024.
Is not an anecdote.
it could be measurement error.
The NAEP only reports "lower" or "higher" for results that are significant.
AI is too new for it
Not everything is AI's fault.
I've known kids
This is an anecdote.
Camera anomalies are not just as credible as drunk anal probing Canadian fraternity wizards riding dragons. Those, plus ducks, balloons and 737s are actually credible.
Lol. People keep repeating this without the slightest idea what it means.
Banks aren't dumb. The restrictions on a billionaire borrowing money are pretty much the same as the restrictions on you borrowing money. It seems like there are fewer restrictions because (1) they're borrowing much less relative to their assets than you are and (2) you probably have only a vague idea of the process based on people making shit up.
In fact, if you're like the majority of Americans and have ever had a car loan or mortgage, you've pulled off the weird trick of using the thing you're about to buy as collateral for the loan you're going to use to buy it. And if you've got a mortgage, you've probably gotten a special sweetheart deal sponsored by massive government intervention.
I'm not sure exactly what you're talking about because you neglected to say what "that" is, but assuming you mean either book or liquidation value, you'd assess SpaceX's book or liquidation value and then give a share of that to Musk.
Yes, it absolutely requires estimation. That's why we have so many different methods despite your assertion that "I haven't seen really any better ideas deployed". The only actual objective method for assessing the value of something is selling it. All other methods are estimates. Market capitalization is arguably the worst because it estimates the value of the whole based on what somebody pays for a (usually) very tiny piece. If I pay you a dollar for a blade of grass on your lawn does that make you a billionaire?
I don't disagree. I said back before the IPO that there were going to be a lot people people buying SpaceX stock as essentially a collectible, like DJT stock. That means most of the value depends on what Musk's next tweet is, or whether he has a heart attack tomorrow.
I don't think it's quite as irrational as you imply though. There is potential in space industry that cheap access opens up. How much is very much a matter of opinion, but it's not zero. Starlink is probably just the first example of something that was not economical at all pre-Falcon but is a pretty good business post-Falcon. There will be a bunch more if Starship delivers, which is looks like it probably will, eventually.
Anyway, "the market" is consenting adults. If somebody wants to blow $170 on a share of SpaceX without any regard for what it's actually worth that's not really much different than spending that much on a $20 bottle of wine at a restaurant. And if you think it's going to be a revolution in whatever, well, lots of people make proportionally bigger bets every day when they decide to start a business.
OS level except *maybe* for the camera
That would break all camera apps that rely on RAW processing, which very much is a thing on phones as well.
RAW, video, anything that grabs stills in real time...
Support bacteria -- it's the only culture some people have!
