Why is a typical human driver at typical performance the baseline? The bad outcomes from driving come from the poor end of the performance, if the average autonomous vehicle avoids those long tail bad outcomes, some inconveniences like slow turns etc is a very worthy tradeoff.

Autonomous vehicles don't get drunk or sleepy, don't speed, don't leave the scene of an accident...

As per Cruise statement:

>>> The initial impact was severe and launched the pedestrian directly in front of the AV. The AV then braked aggressively to minimize the impact.

The target market is enterprise. If I showed up for my first day of work and they handed me a 2019 MacBook Air I'd be pretty unhappy.

I really don't get this attitude of indignation when a powerful machine's resources are fully utilized. What else are you doing with with that computing power? There's not some finite pool of computing power we are depleting here, they are figuring out how to make more all the time!

It's your right as a consumer to have a pro-consumption attitude, because you're spending your own money. Be thrifty. Be profligate. You choose.

But having it as a developer is a different matter. That "powerful machine's resources" aren't yours. They belong to your customer, and it isn't your business what your customer is doing with the computing power your program isn't using.

Writing programs that demand more resources than necessary forces people to discard and replace otherwise perfectly good computers. Which, again, you as the developer don't have to pay for.

If tragedies of the commons don't bug you, then none of this matters. But if they do, then imagine millions of people ordering millions of cardboard boxes from Amazon containing millions of sticks of DRAM to run that O(n^4) loop that you're about to check in.

I can sympathize with this attitude with things like a Slack clients, which are designed to run in the background, and which developers probably don’t have much choice over.

But there are plenty of lightweight IDE options. IntelliJ competes to be the most powerful IDE available. That implies using lots of compute resources for productivity gains that might be marginal or not even realized by all devs.

Equality of access to compute around the world could be a principle pointing towards doing more with less.

The 2019 Air is underpowered (1.6 GHz Dual-Core, 7W TDP[1]) even when comparing it to more affordable hardware from 2019 or secondary market business laptops from the prior 5 years or so.

[1]: https://www.intel.com/content/www/us/en/products/sku/189912/...

The problem isn't that people are hogging all the CPU, the problem is that it's hard to get computers to people around the world, connect them to infrastructure (power, internet), and train people to use them. Plenty of smart people have actively tried to solve this problem and have failed. I'm not sure if adopting such a principle means anything whatsoever in the face of what appears to be a larger challenge than providing computing power to a willing yet deprived population.

Twitter was a second tier advertising destination before there was any inkling of a Musk deal, and advertisers started leaving before the deal closed when he was trying to get out of it. The biggest advertising firms in the world are advising their clients to leave.

Why would they come back? Musk is the Donald Trump of tech - plenty of devoted fans, but not someone brands want to associate themselves with. Even if he wanted to, it doesn't seem that Musk can stop impulsively tweeting controversial things.

The advertiser exodus (already well underway) is a far bigger threat to Twitter than a user exodus. The subscription model is dead on arrival.

There's plenty of options if you want a forum without "corporate ESG posing" - 4chan et al. They just don't make money, or interest the majority of people.

How does Binance fit into this chart? I understand them to be the biggest remaining exchange

You couldn't previously buy a checkmark, now you can. This opens new methods of attacking the network.

This move perhaps limits spam as practiced today, but the attacks that will happen once the network changes will be different in ways that are difficult to predict.

It's a great point. There have been several replies ITT along the lines of, "if Twitter goes away, what platform will Stephen King post his content?"

As you point out, the answer is "all of the relevant ones, with very little effort on his part"

The critical thing for the platform is that if I want to find the Twitter account for {celebrity foo} I can do so with a high degree of confidence that it will be real.

I think the loss of trust from consumers is the bigger risk, successful impersonations are relatively high profile and people don't like being tricked.

There's no meaningful difference between what you are describing and advertising.

Advertising is a great business to be in apparently, maybe he's diversifying from corporate clients.