We would certainly be at a disadvantage if Russia was able to copy rockets that constantly blew up shortly after launch.

Yeah, using Intellij via Amazon Q. In "theory" it should've been on the same context but maybe there's some quirk somewhere.

There was a separate time where I had a really complex test, same code, that required merging 4 different data sets into a unified data set. I generated the 4 different test sets from real world data then gave the prompt about generating a test for the merge using these 4 different data samples and... it, first time, made a proper test with mocks that read the files at the appropriate time and then verified that the returned, merged result was correct.

It's just wild that that complex unit test was 100% on the first try but simpler tests went off into the woods. But, like I said, that's been my experience with AI, so far.

I'm not sure that's true, at least for Claude Sonnet 4.5. In the same chat I had it write up some unit tests that exercised a function and verified that certain external libraries were called via mock. It wrote up the tests and they looked good but the mock syntax was incorrect. So, in the same chat, I pointed out the error and asked it to fix the syntax issues. It churned for a few minutes and couldn't figure out how to resolve the issue so it decided the best solution was to simplify the test... which it did... and removed all the mock verification that the external libraries were called! So I pointed out that the test looked good but still had to verify that external libraries were called via mock. To which it responded, yes I can do that - and proceeded to generate a wholly different set of mocks, perfectly.

That's been my experience overall so far. Sometimes the AI gives a bang-on solution and code and sometimes it just flails about. I'm sure part of that is how I'm prompting but even then the AI seems to respond differently even if I phrase/word instructions in a similar fashion.

Ya gotta go all Captain Kirk on that AI!
https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.youtube.com%2Fwatch%3F...

I used to work for Sling TV, and you basically have that backwards. ESPN is the part of Disney's package that people are willing to pay money for. The shutdown and negotiations every year is just Disney forcing the various providers to pay for and carry their other channels. That's why Disney always holds these negotiations during football season, so if they have to shut someone down their customers actually care. Every year viewership on Disney's other channels (and non-sports channels in general) is lower, and the prices that the content producers require goes up. Scripted television is in serious decline, and Hollywood is using sports fans to prop it up.

As an example, If you don't care about sports you can get Disney+ without ads for about $12 a month. Disney will happily throw in Hulu for that same price if you will watch some ads. You can binge watch the shows that you care about and then switch to another channel. Heck, you can buy entire seasons of their shows ala carte. You can't get ESPN however, without paying at least $45/month, and that's with a package with no non-Disney channels and chuck full of ads. For the record, that's basically what the streaming services are paying Disney as well. When I worked at Sling the entirety of the subscription fees went to the content companies (primarily Disney). There is essentially no profit in cable packages. All of the profit has to be made up somewhere else.

People that aren't sports fans, especially if they are entertainment fans, tend to believe that scripted programming is carrying sports, but it is the other way around. That's why AppleTV, which has spent over $20 billion creating content for their channel has about as many subscribers the amount of people that typically watch a single episode of Thursday Night Football, the worst professional football game of the week. Amazon Prime pays $1 billion a year for that franchise, and it is a bargain compared to creating scripted content. Apple makes great television that almost no one pays for. The other content providers are in the same boat. You'll notice, for example, that Netflix's most expensive package is $25/month, and the revenue per user in the U.S. is around $16. That's ad free. The lowest promotional price you can pay for ESPN is basically twice that, and it always comes with ads. What's more, sports fans tend to actually watch the ads.

Sling is selling day and weekend passes to people because it knows that most of its customers only have their service to watch the game. No one is watching linear television anymore, but the content creators have built their entire business around the idea of having a channel that they fill up with content. Even with Sling's ridiculous prices they can typically watch the games they want to watch for less than maintaining a subscription.

I have spent most of my adult life in the sports world, but I don't watch sports. I personally believe that in the long run sports television is probably going to end up uncoupled from scripted television. I think that is going to be very bad news for people that like scripted television.

Anthropic's entire pitch has always been safety. Innovation like this tends to favor a very few companies, and it leaves behind a whole pile of losers that also had to spend ridiculous amounts of capital in the hopes of catching the next wave. If you bet on the winning company you make a pile of money, if you pick one of the losers then the capital you invested evaporates. Anthropic has positioned itself as OpenAI, except with safeguards, and that could very well be the formula that wins the jackpot. Historically, litigation and government sponsorship have been instrumental in picking winners.

However, as things currently stand, Anthropic is unlikely to win on technical merits over its competition. So Dario's entire job as a CEO is basically to get the government involved. If he can create enough doubt about the people that are currently making decisions in AI circles that the government gets involved, either directly through government investment, or indirectly through legislation, then his firm has a chance at grabbing the brass ring. That's not to say that he is wrong, he might even be sincere. It is just that it isn't surprising that his pitch is that AI has the potential to be wildly dangerous and we need to think about safety. That's essentially the only path that makes his firm a viable long term player.

He used to win these market timing games because no one was paying attention to huge short positions. You could quietly bet against a company, or, better yet, you could quietly amass a short position and then release stunning negative news that you had uncovered and watch the stock price tank.

These days it is more likely that online investors will notice a large short, and drive the price of the stock up until the person holding the short gets margin called and loses all of their money. The shorters then provide the liquidity you need to get out of the position. There used to be good money in shorting terrible companies, but in an age where hordes of armchair investors can drive the price of GameStop to the moon that strategy is just too risky.

I suspect that these streams are costing Youtube money. They can't monetize them, and they have to spend effort shutting them down or they get in serious legal trouble. If things persist I suspect that Youtube changes its live streaming service in ways that make this impossible.

We will see though.

The problem, of course, is that Sports content is paying more than its fair share of the bill for all televised content. It is easy to see the large bills and assume that sports is a cost center, but the reality is that sport tends to pay its own way, while scripted television is much more of a gamble. To a certain extent that is why most scripted television these days is so formulaic. The television studios know that they can make money with modern versions of "The Rockford Files." That's why NCIS is in its quadzillionth season.

Severance is great, but it is a prime example of what I am talking about. Apple has spent billions of dollars on content at this point, and they are still hemorrhaging money. People like their shows, but they aren't lining up to pay for them. Shoresy is in a better spot, but only because Disney is doing its level best to tie Shoresy to ESPN and other sports related content that people are willing to pay for. The folks wanting to buy ESPN can get the rest of the Disney bundle for pennies. You can't just buy ESPN, you have to buy it with a television package. Disney does this because they know that if people have their other channels, then they tend to watch them. They are willing to pay a premium, however, for sports.

Hulu is cheap, and you can get it by itself. The same goes for AppleTV. All of these cost Netflix amounts of money $12 (or so) a month. When I worked for Sling it's entire packaging was based around making it possible to bundle ESPN for less than anyone else. If you want ESPN the least you can pay is $45/month, and that doesn't give you the other channel's sports package, that you probably want if you are a sports fan as well. It is very likely that the team that you follow will have at least one game on ESPN's competitors. That means that if you are purchasing from Sling you need the blue package as well (which is another $45, or bundled will total $60). You could easily sign up for all of the non-sports streaming channels for less than an Orange+Blue package (which once again is as competitively priced as it is possible to do). I was just looking at Disney's bundle, and you can get Disney+, Hulu, and ESPN for $35/month, which is definitely the least expensive way to get ESPN these days. That's with ads, which are added even to VOD content. If you want to watch your VOD content without ads that's another $10. Linear content (like watching cable) always comes with ads. Sports fans can't dodge ads ever.

I bring up pricing like this to make it clear which parts of television customers are actually willing to pay money for. If you don't want to pay for sports (and I don't blame you), then you can easily pay $12/month and switch between streaming providers and watch whatever shows you want to watch. All of those services allow you to easily stop and continue your subscription, and none of the content is likely to go away. Heck, chances are good that, if you wait long enough, you can watch the shows that you want on one of the free services. In most cases they are literally giving away old scripted content. The problem with this model, is that it doesn't make Hollywood enough money to be profitable with their current structure. The reason that Disney (and everyone else) bundle channels the way that they do is because they know that they can't afford to gamble on scripted content unless they bundle those risks with the proven money generation of sports content. More and more people like you, who don't want to pay for sports content, are opting for less expensive alternatives that still get them the shows that they want.

This market contraction is why Hollywood is so focused on franchises that have historically been popular. So instead of new shows we get derivatives of things that were popular in the past. Scripted content is risky, and as it gets uncoupled from less risky sports content producers do whatever they can to hedge their bets. So we get a re-re-remake of the TMNTs, Spiderman, or we get another cop show. Recently we have also been blessed with shows that have been popular in other countries or markets (that is legitimately cool in my opinion), but that is also likely to dry up as entertainment becomes more global.

Which leaves what can be done on Youtube budgets for anything remotely risky. Which is fine, I suppose. Personally, I like watching people restore old sailboats. That's not something that is ever going to be more than a niche market, but on Youtube that's enough of a market to make it financially viable for a few people. Maybe with AI it will even become possible to do good SciFi with that sort of a budget. Who knows? One thing is certain, it is definitely interesting times ahead.

Plus, IBM owns RedHat. So that's probably something. Then again, maybe that is the hyper growth software and services bit that they want to keep.

A direct line between County Cork and Loudon VA does not go through the east coast of Maryland on a spherical earth, that was the only point I addressed. Even if you wanted to maximize how much is laid in the ocean as opposed to land, there are still shorter distances that would land you on at least in Delaware. While I am sure there are logistical reasons to do it where they are doing, that is irrelevant to my point.

Yeah that sounds off. Probably mean digging at coastal shallower waters then just having it lay on the floor elsewhere, that's the standard.

That's assuming we live on a flat earth. A shorter line between County Cork and the US would have the US terminus much further north.

"We are crazy."

What's incredibly frustrating is how much of the craziness has been created and driven by Australians. Other countries want to push back against Trump? Go after the lunatics who fill his dementia-ridden head with all the crazy ideas, including their companies.

My hyperbole got the better of me again. You are not the only one. Little House on the Prairie is remarkably good television. It definitely beats doom-scrolling on your phone.