infamia

> No they weren't. If this administration had strong reasons to believe Microsoft cheated and engaged in fraudulent practices, the DoJ would have opened an investigation leading to punitive measures. It's a dog whistle and red meat for "the base."

Microsoft and Adobe are both already under investigation. Based on the preliminary findings, they've been suspended from participating in the Green Card program along with a host of other companies. [1]

> That's not going to happen. It's easier for them to send the jobs offshore.

It sounds like that's what the administration is demanding (i.e., hire more US workers before they can participate in the H1B program). [2]

[1] https://www.cnbc.com/2026/10/08/microsoft-adobe-green-card-l...

[2] https://www.theguardian.com/technology/2026/oct/08/jd-vance-...

> This is the problem I have. People are built to be fooled by this stuff - to see something that's not there. Pareidolia but with language.

I really like this metaphor, it hits home that just because you sense something is human-like doesn't make it so. We're wired to respond to those sorts of things.

> When I stumble around trying to explain my thinking

I had similar difficulties expressing my thoughts in an organized way. The video below from Richard Sutton (the father of Reinforcement Learning [which LLMs use extensively in training]), helped bring order and firm up some of my thoughts and theories.

https://www.youtube.com/watch?v=21EYKqUsPfg

>I invariably get hit with the response, Sure, but if it's functionally identical to a coin being magically pulled out your ear, what's the difference if it is or if it isn't?

Not to spoil the video, but it isn't at all functionally equivalent. LLMs are an imperfect and somewhat randomized simulation of what the LLM thinks an average person might say in response to a question. Which is another way of saying, most people will on average get worse answers than if they worked on a problem themselves (but they will get that worse answer from the LLM comparatively quickly). Practical people respond to this positioning fairly well, and for others it still lands a bit since no one wants to think they're below average. That said, I'm still working on my positioning a bit as well!

And a magician making a coin "disappear" doesn't mean that magic is real. I see no way we can call it thinking without a goal (other than computing the next token).

It sounds like your app is using a lot of dynamic techniques like meta-programming? If so, that makes a lot of sense. Most of the apps I've seen use very little use of meta-programming and friends and incremental typing worked rather well.

Unless you're doing something really dynamic, I don't really understand this position and would like to learn more. for most cases gradual typing is a great option that can be done incrementally in Python (two things even cheap agents are really good at).

If you're looking for "Hot Dog/Not A Hotdog" type (i.e., heavily bounded) answers from LLMs, make certain to at least consider (or better still benchmark) traditional AI and machine learning approaches against LLMs. The traditional approach is almost certainly less computationally expensive and might work better for your problem than an LLM. I recall hearing stories from people on a popular podcast, that they benchmarked and went with the traditional approach because it was more accurate, deterministic, and cheaper. Many had always wanted to try traditional AI/ML, but only got the go ahead recently to add AI to their project from management because of the AI/LLM hype. They were surprised to find that when they benchmarked, the traditional approaches outperformed LLMs both in terms of accuracy and cost. Don't get fixated on LLMs to the point of ignoring older tools that might do a better job.

No one knows how many bugs claude has because Anthropic auto-closes Issues (or at least used to) if there isn't someone constantly pinging the issue every two weeks. I've contributed to several issues that were confirmed by several other people, which were they were auto-closed after people gave up confirming the problem without any response from Anthropic. They seemingly don't care if it isn't on fire or at least smoldering heavily, which seems really bad in my book if you're trying to make even reasonably good software.

> Heavens. Is it possible you're reading into it a bit?

Perhaps, which is why I hedged with the possibility that it is mere hyperbole. However, I think what did it for me was his statement that he provides and expects feedback to anyone not following this recipe, which is at utterly without nuance. Incidentally, you can execute with focus and intention without urgency and arguably be more effective over the long haul.

My point is that after a large change, reversion to the mean should not be treated breathlessly, but viewed in a measured manner.

> Before the 2024 Carbon Pollution Standards, the federal limits were much weaker—and for existing power plants, there was effectively no operative federal CO₂ limit at all.

This is an exaggeration that stretches credulity, since there have been strict limits for decades around secondary outputs that are required by power plant energy production effectively limiting production and emissions (which is the entire point).

> Something that people learn quickly when they work with me is that my approach to pretty much every problem is: > ... > 6. Act with urgency to achieve the goal

If everything is urgent, then nothing is. This guy sounds miserable to work for and with. Assuming this is accurate and not just hyperbole, he is essentially saying he has no prioritization skills because everything is urgent. I think most people who have been around the block have worked with people like this, and unbeknownst to them, their coworkers develop a default snooze button associated with most of their requests and projects.

This headline is rather sweeping considering the modest nature of the change, which is merely rolling back regulations to 2024 levels.

It says they offer a degoogled version, which is cool.

https://minimalcompany.com/#purchase

In the long run a planned economy is a huge disadvantage, the Soviets and now the the CCP are prime examples. For instance, the CCP is also sitting on a bunch of practically empty metropolises, paired with a property bubble and a falling population (all of which are exist to their meddling).

> The ADA isn't some surprise gotcha that crept up on any of these people

The ADA was written in an extremely non-specific way that allows unelected bureaucrats to reinterpret it every so often. To be honest, the US would probably have a more accessible society if the law was better written.

> I do not usually have every word I will respond with stored in my limited working memory. My output is often predicted based on the previous word I spoke.

People don't know exactly the words that they're going to say necessarily, but tend to start with a general concept of what they're trying to communicate and only then try to put together the words (sometimes out of order). LLMs do not begin with any sort of concept they're trying to express. LLMs are simulations that attempt to reproduce what an average person might say while wired up to a huge knowledgebase.

Great point! I'd also add that Django's another solid choice for boring tech that LLM agents will know very well and will very likely continue to do so. A fair bit of SWE Bench and other Python benchmarks are Django related tests, which the LLM vendors care very deeply about keeping their scores up. Also, Django's docs are excellent, so strategically pointing an LLM to them in a prompt can often produce great results.

If you want to just prefetch everything throughout a project or just on a per-Queryset basis, all that is coming in the next release of Django.

https://docs.djangoproject.com/en/dev/releases/6.1/#model-fi...

That's the great thing about Django, it's been around so long and the quality bar is so high that eventually all the major rough edges get sanded away usually in a really well considered manner.

That's a really cool story! I live near a couple of industrial sites like that which have a power plant on site, that were shut down quite a long time ago due to environmental concerns they don't burn cleanly at all as you said!

A transmitter is a sensor in a power plant that very precisely measures the air flow within the combustion chamber, so the fuel burns as clean/efficiently as possible. Relatively new transmitters can read with more precision, so you can dial down the airflow/fuel and the corresponding power output.

> That depends on the design. ... The older the design the less likely it can turn down far.

I don't doubt there's a small number of plants that can't do 10-1. That said, transmitters are intended to be replaced and upgraded regularly since they have a finite lifespan (they're cheap anyhow). 10-1 has been around for 35+ years now and you're really lucky to get 20 years out of a transmitter.

> it's logically incoherent to try to determine what level of inflexible power generation will meet "the baseload".

Lots of folks don't know that coal and gas fired plants are able to turn down their power generation more than 90% (i.e., "ten to one turn down"). It's definitely true that cold restarts take a while. However, they often rather quickly and dramatically change their power output profile while running.

> In fact, the complexity of Chat Messages and Notifications synchronisation is one of the prime reasons why React exists.

SSE is perfect for chat/notifications and is a lot simpler than syncing. The server pushes updates to the client doesn't require a sync. SSE is also more performant since it can keep the connection open. The example below is on a $10 vps last I heard.

https://data-star.dev/examples/dbmon

> Having control of something isn't the same as third party services granting you access. Many online games only supports windows with secureboot and TPM enabled, but that would be a silly excuse to say that you don't really "own" your PC.

Your example is about one developer's decision, which is not really what we're talking about. The damage goes way up when you start talking about an arbitrary number of applications. A more apt analogy is, what if you couldn't install any applications except through the Windows Store? We're talking about restrictions laid down by the OS developer and for iding all applications outside of their crappy walled garden, which is very different. Also, you don't actually think Google is going to stop there do you? They will certainly turn it off one day of they think they can get away with it, because it is in their financial self-interests to do so.

A bit more like Kevin Powell, who publishes lots of free CSS videos [0], has some good basic CSS Training Courses for free, but charges money for the advanced ones [1]. His videos are not only available on his site, but they're also on Front End Masters, which is a well known training site.

[0] https://www.youtube.com/@KevinPowell

[1] https://www.kevinpowell.co/courses/

I think for most people you're right, they just want to upload their videos, and maybe make a couple of bucks on the side. I was referring to folks who are truly committed to making this their career and want to have their own brand. Many, many people are already doing this, so it isn't something theoretical. Most YouTubers are automating their production pipeline, so another upload step isn't too hard, especially nowadays where it is easier than ever to build a bespoke, deterministic pipeline with agents writing scripts and programs.

There are lots of strategies how to make money and it will vary based on your specialty. It could be premium content, premium services, exclusive engagement with your fans, merchandising (e.g., many cooking YouTubers sell their own brand of cooking utensils that they have designed themselves). It requires some thought, but the idea is to deepen the authenticity and engagement by building an actual community and your own personal micro-brand. The offering needs to feel organic and huge value to your community in some way. Many many people have done this on YT and now are making way more money than what's possible with the pittance that is YT monetization. YT then becomes just one part of your funnel intake strategy once you get big enough.

Edit: Of course, you do need to have enough demand/scale to make it worth your while, but that will depend on your audience size and how engaged/invested they are in your content AND you personally to an extent. Perhaps be creative and start out with small experiments. Not too hard with LLMs nowadays.

> Realistically, how many viewers will be retained should YT shut the OP down? Right now, that number rounds to 0. Practically speaking, YT is free internet video streaming for long-form videos on the US market.

People have been following this strategy of creating their own mini-brands for years and have their own following. YT doesn't even have all of the their content, and YT is just one lead gen channel. Frankly, Ad Revenue from YT is pitiful, and it's an open secret the real money is made as I've described (although it is a long term play).

I don't disagree with you about PeerTube and the like, but it is a two-sided marketplace and you need to prime both sides of that pump (content creators and viewers).

You can publish to both and even better your own domain that simply points to your video hosting provider. Long term you want to own your distribution channel as much as possible, while using YouTube as your lead generation tool to drive true believers to your site and premium distribution channel not owned by YouTube. Otherwise, you will always be subject to platform risk via YouTube's whims which has destroyed many content creators. That's the long term winning play IMO and it doesn't preclude tools like FreeTube.

> Requiring a disclaimer is essentially admitting the content isn’t meaningfully different than human generated content. At that point, who cares? Just engage with the premise on its own merits, rather than on how it was written.

The problem is the reader has to invest time to find out and LLM written text will (on average) lower the quality towards "meh" and spend more words doing so. Even if the author is making an earnest effort to produce high quality content, they need to admit to themselves and others that their results will be more hit or miss. The disclosure allows the reader to make a more informed decision about how to engage with the material (e.g., have an LLM summarize or analyze the content, or just dive in because we know it will very likely be a good read). Editing what someone has written is like reviewing code, you're by default not as invested, so the results will likely reflect that reality.

Great advice! The really tricky part to me is when you have an event your recorded before it happened, but want to look it up after the event has passed (e.g., you want to look up the doctor's appointment a year after it occurred). The simplest and mostly solid answer I've been able to come up with is:

1. If you want to know when something happened and a particular place is important (like the previously mentioned doctor's appointment), store the local date/time with timezone data. That covers you in case the timezone changes before your recorded event happens. Personally, I would not store reflexively store dates/times in a string. For the cases I encounter, that feels like primitive obsession since you can always use EXTRACT in a query to simplify output.

2. If will you need to lookup the date and time after an event occurred, write a separate field that includes timezone offset field (e.g. -1, +1, -8, etc.) in case you want to look up exactly when an event previously happened. This (mostly) covers you from timezone shifts that occur at that particular location between when you wrote the data vs some later date. This falls apart if the timezone you're converting to also changed their timezone between now and the event. Also, if your timezone shifts between when you wrote the record and the actual event.

I wonder if someone has a temporal record of timezone shifts. You could solve a lot of edge cases with something like that. Then you could write a query that asks for the timezone's offset as of a specific date. That would make life much easier. Then you could skip the timezone offset field I mentioned in #2.

It's no huge secret that most large, modern countries are researching vaccines for these dangerous diseases. There are quite a few of these labs on U.S. soil as well last I checked. I don't get what the scandal is supposed to be about, other than ensuring the labs remain safe during the war?