He's running the same playbook. Threaten to tariff/investigate/restrict/ban an industry and then wait for that industry to pay him off.

Given how AI behavior can be unpredictable or unexpectedly stupid, what are the odds that someone created an AI specifically tasked to look for and take advantage of any weaknesses or shortcomings in the hedge fund AI?

This is just more bluster from russia; simply another desperate attempt to get the rest of the world to pretend russia is a power to be reckoned with. Their economy is collapsing due to their war against Ukraine. Their only spaceport is damaged and unusable. They're selling themselves off to China and becoming a resource colony. They're selling their oil at a loss. russia has become a joke under Putin's rule.

How are they planning on carrying over code comments to the new code base...?

For example, the Target app charged people $100 more for a t.v. when a person was in a Target parking lot versus when in another location because it appeared to determine people would pay higher prices the closer they were to the store.

https://consumerwatchdog.org/privacy/new-report-details-how-companies-use-surveillance-to-charge-different-prices-for-the-same-item/

Amazon is "pledging" to expand capacity for government contracts? Any reason why we're not using the phrase "going to bill the government to cover the costs of the expansion" instead?

What's the point of freezing a body? The water in the cells freezes, expands, and ruptures the cell membrane. The body is effectively mush at that point. Does anyone really expect medical tech to get to the point that it can repair that kind of cell damage in every cell in the body?

LLMs are randomized algorithms. They return a random value from its weighted set of dialogue options. As such, I'm not sure I would put LLMs in the same category as AGIs.The article below is interesting about how non-determinism works in such algorithms and why such randomness is actually useful.
https://towardsdatascience.com/llms-are-randomized-algorithms/

Of course this means, at best, I will never trust LLMs to be anything more than a convenient but lazy data miner, never mind that I'll still need to double check its answers/sources.

Let me guess, this story was submitted by Apple's marketing department?

Steam isn't an evil monopoly, instead I would argue it is a "natural" monopoly because they provide an excellent service without treating their customers like cash cows to constantly butcher for an additional buck, unlike some other competitors and businesses in general.

If you can force a company to agree to and normalize such winks, then the next step is to expand the protocol to communicate additional information. It's basically a step towards forcing companies to provide intelligence.

Yup. You can either pay the full 100k to the US Government or instead pay a portion of that 100k (or an equivalent favor/gift) to Trump.

This is the problem with a single idiot making up the rules on the fly. You can't trust the rules won't change based on the last person he talked to, the last bribe he received, the latest opinion polls, or whether his drug regime was changed.

Another reason for having actual laws passed by Congress* is that Congress is slow to change them and doesn't like to do so willy-nilly.

* Not to imply that the GOP Congress is actually capable of passing laws or pushing back on the Executive or standing up for America in general.

So, training LLMs on the sum of human knowledge wasn't enough to make a human equivalent AI? The human brain runs on ~20 watts of power; OpenAI/Oracle/SoftBank are building five data centers that will use ~10-gigawatts of power, which is ~500 million of those 20 watt human brains. Guess it's back to the drawing board on how to create a human level AI?

The "common people" are going to learn about the importance of testing the hard way...

Microsoft says its Agent Mode in Excel has an accuracy rate of 57.2 percent in SpreadsheetBench, a benchmark for evaluating an AI model’s ability to edit real world spreadsheets. This result places Agent Mode above Shortcut.ai, ChatGPT agent with .xlsx support, and Claude Files Opus 4.1. It’s still behind the human accuracy of 71.3 percent, though.