RoddaWallPro

Missing context from my comment: I think the triple IPOs of SpaceX, OpenAI, and Anthropic, are going to be the thing that pops the AI bubble. They IPO for X, the stocks fall by some large double-digit percentage on future news (some model training didnt go well, or maybe China releases another DeepSeek model that tanks Nvidia stock like prior, etc), and the whole tech industry public market goes south. And since, from what I understand, most of the gains in the last few years on the stock market have been driven by AI-inflated valuations on a few large players, that crashes the whole market, and then we're _really_ in pain.

I don't know if this is realistic. But I look at these numbers and its like something does not add up here. The numbers are _way_ too high. As Altman said, "someone is going to lose a phenomenal amount of money". I don't want to bet on a specific player, because I don't know who, but I want to bet in general that someone is going to lose their shirt.

I want to buy options against QQQ so badly -- but Tesla has traded at a crazy multiple of revenue/profits for a very long time, so I'm wondering if Elon/AI hype will keep these stocks high longer than I want to pay the risk premium for (options).

I have a really hard time with this argument because I'm _positive_ 99.99% of bullets fired in the US are NOT being fired to kill things. So I see people this arguments and its like, hm, interesting. Interesting that the overwhelming vast majority of the use of this thing is NOT the use that you are claiming it is used for. Doesn't hold up.

Mitchel Hashimoto started a tool for this, https://github.com/mitchellh/vouch.

I'm saying it _is_ and _was_ an issue during the day and heavy commute hours, those were the only hours I rode it! Other places in the world with nice train systems do not burden their riders with "safety in numbers", the places are just plain safer, period. And a great place to start is Don't let people smoke fentanyl on the train :) (And make sure everyone has affordable housing and healthcare, ofc)

I rode the MAX when living there for a few years. I vividly recall screaming drugged out homeless riders being a regular feature. The last time I rode, a year ago, there was someone in the throes of the fent-bends in my section, who smelled like he was dying (he well may have been).

These incidents haven't made me fear, because I am a relatively big and tall male, but they _definitely_ will for others. And even then, they aren't pleasant.

You simply don't run into those things often on trains/subways in Europe (I lived in Spain for a year and traveled extensively in Europe during that time, and on other europe trips prior). So fix those issues, and then I am sure people will want to ride the rails.

5 years ago was the beginning of 2021, just under a year after GPT3 was released (which was not good at doing anything useful). And that model was 175B params.

GPT4 has been widely rumored to have 1.8 trillion params, which is 10x more, and was released 2 years after this "5 years ago" date that you are using here.

So, to quote yourself here, "This is not true and unfortunately this significantly reduced the credibility of this article for me" /s/article/comment

Actually, that is not what is happening here. What is happening here is that the govt is saying "Okay, we will not buy your widgets. Also, anyone who _does_ buy your widgets, regardless of what they are doing with them, we the government will not do any business with them." Which is waayyyy beyond just not buying widgets. That is outright retaliation and using your power to attempt to destroy a company.

A 30 day waiting period on news articles doesn't meaningfully reduce actual suicides. One on guns _does_, without a corresponding harm to the buyer.

What do you mean exactly by "diversify"? Money/investment-wise?

I do something similar - I create a "2025December.md" file each month (with proper year/month obviously) and have a bullet list of everything I'm working on/trying to keep track of. I also use it as a scratchpad for whatever, and writing down notes for projects. Each day I insert a "#### 11 Dec 2025" heading at the bottom of the file, then just copy over everything relevant from the previous entry.

It's stored in my Dropbox so it is always backed up, though it is not VCS'd. It's worked for me for years, far better than any app. Too, I have full control over it, and years of the data, free for processing by any tools/LLMs that I might want (I haven't wanted such a thing so far, but maybe I will).

I _kind of_ understand this one. You can think of a bubble as a market exploring a bunch of different possibilities, a lot of which may not work out. But the ones that do work out, they may go on to be foundational. Sort of like startups: you bet that most of them will fail, but that's okay, you're making bets!

The difference of course is that when a startup goes out of business, it's fine (from my perspective) because it was probably all VC money anyway and so it doesn't cause much damage, whereas the entire economy bubble popping causes a lot of damage.

I don't know that he's arguing that they are good, but rather that _some_ kinds of bubbles can have a lot of positive effects.

Maybe he's doing the same thing here, I don't know. I see the words "advertising would make X Product better" and I stop reading. Perhaps I am blindly following my own ideology here :shrug:.

I agree that is what he is doing, but I can also justify adding fentanyl to every drug sold in the world as "making it better" from a business perspective, because it is addictive. Anyone who ignores the moral or ethical angle on decisions, I cannot take seriously. It's like saying that Maximizing Shareholder Value is always the right thing to do. No, it isn't. So don't say stupid shit like that, be a human being and use your brain and capacity to look at things and analyze "is this good for human society?".

"advertising would make ChatGPT a better product."

And with that, I will never read anything this guy writes again :)

But just to clarify, because I'm also having a hard time imagining this, an LTE antenna in a cell phone can beam data to a satellite and have it picked up? Even at whatever low kbps? That is insane to me!

With (I assume) cell phone use prevalent in every single classroom in the nation that hasn't banned them, and school shootings a minuscule probability, "much rarer" is doing a lot of work here hah.

https://marketplace.visualstudio.com/items?itemName=stkb.rew...

I use this extension extensively. It's not auto-wrapping, but you can bind it to an easy shortcut and wrap when you need to. I find it almost indispensable. I wrote a vscode extension to do the same thing, then discovered this one which does it far better.

I have what is probably a dumb question: How can a Raptor turbopump need almost double the HP of a F1 while putting out 1/3 or 1/4th the thrust? (Assuming Elon's 100k HP number was correct, and/or hasn't changed). That just doesn't settle out for me. If it's got double the power, it should be moving double the fuel, so double the power, no?

I had a subscription to the WSJ for awhile. I liked it because it is consistent. It definitely has a conservative slant, especially in the opinion section, but you can "lead the shot" so to speak. You can account for it. It is reliably and consistently conservative, and being conservative is relatively consistent.

The NYT or Atlantic or New Yorker, etc? God only knows what new thing has been declared off-limits/"problematic" in the online progressive world this week, and the tone of self-righteous goodness from the progressive media is insufferable.

I feel like Matt Stone from South Park: "I hate conservatives, but I really fucking hate liberals."

"shift/control-clicking to get multiple things is a flat no-go from a UX perspective" - Do you mean that this is NOT how you should do multiselects? If that is what you mean, then how _do_ you do them? If I have a list of items and I want to select 10 or 15 of them in a row, I currently don't know of a better UI to do that with than shift+click.

As someone who also lives in Central Oregon (and was born and raised here), I will say this: this article is profoundly misleading.

They state that the Upper Deschutes CFS goes from ~1200->~65, and state that is from irrigation. But they don't actually prove that. Much of that CFS decline could be observed out of irrigation season, because that section of the Deschutes dumps a ton of water into lava tubes/back into the ground. You'd have to measure the CFS at the end of that section prior to irrigation getting turned on, then after, and then the difference is what is going to irrigation. And CO Land Watch doesn't do that. Hell, they don't even state where their data is coming from.

You can get a better idea of this drop by instead looking at official US govt data where the Deschutes drops into Lake Billy Chinook: https://nwis.waterdata.usgs.gov/usa/nwis/uv/?cb_00060=on&for...

That is for the year 2021, when that article was written. You can see the drop in April, which is when irrigation starts. It's about 500CFS,or ~half the river at that point. That is NOT 95% of the river, as CO Land Watch would have you believe. (Side note: CO Land Watch has a bit of a reputation around here. This is a bit of an exaggeration, but it seems like they would gladly end all farming in Central Oregon to save a few salmon.)

About every quarter, I get together with a group of friends a few hours away and we have an all-weekend LAN party. They all live in the same area, and have wives & kids so usually people are in and out as their schedule permits, but we all hole in up a house and play games for a few days. It's fantastic.

It's also allowed me to completely deal with any video game addict urges that I get. If I feel the urge to start playing some recently released game, I just tell myself "Ah, I'll play it at LAN", and then I don't need to play in the moment. The urge goes away. So now I only play 2 days a quarter at LANs (maybe 8 days a year total), whereas previously years of my life would go into games. It's been a hugely positive thing for me.

This is a nitpick, but their "Claude 2 on 200k Context Data" graph doesn't actually extend to 200k, only 100k. Would be curious to know if that _is_ actually the graph for 200k with the wrong axis, or if its the 100k graph.

I've been wondering if all of the jailbreak-fixing/rlhf-tuning that is happening to GPT4 is responsible for "nerfing it" (Still unsure if that's actually happening or if people are just noticing the gaps in its understanding more now).

Imagine someone who is perfectly politically-correct and never says anything even remotely edgy/original. When I imagine people like this (who I've met irl), they are genuinely a little bit stupid. And I wonder if the "make this model never output anything "dangerous" process" causes a model to become stupider.

Anyway, I'm off to go see if Claude 2 will help me stage a coup in a third-world country and become its dictator. Adieu.

I relate to this, though perhaps obliquely.

I feel far more peaceful and centered when I am hungover. I don't have a thousand things flying around the edges of my brain, nipping at my consciousness. I just _am_.

Drinking can do that for me too, but it shuts off so much of my brain that eventually, it's like being what I imagine a dog is. Completely reactive, with almost no higher thinking. Not that that is bad. I enjoy it.

If not for how bad it makes me feel physically, I'd probably spend most of my waking hours either drunk or hungover. What that says about me, I hesitate to think about, hah.

I also found that in my limited experience with meditation, I could sometimes find that place of tranquility and "simply being". But it was so rare, and the experience of meditation so frustrating and seemingly impossible to progress in, that I gave up. But I still hold that what I am after in meditating is the same thing that I am after in a night of drinking. Single-threadedness, peace, acceptance.

I requested & have been waiting for access to Claude for nearly 3 months now. Guess the waitlist must be really long...

Fool me... you can't get fooled again.

George W gaffes have become so hilarious to me, now that they're 15+ years in the past. I laughed out loud reading this one.

Yeah, that's probably true. He also at least allows people to set their own reading width by adjusting browser.

My frustration stems from the fact that I find the argument "there are no papers with sufficient evidence" to be pedantic bullshit. Like yeah, sure, you aren't even wrong, but absence of evidence is not evidence of absence. I've never seen anyone claim to like 180 char lines, whereas I've seen hordes of people who say it is very difficult for them to read that line length, and prefer something book-sized (lengthed?).

One of the more hilarious takes I've seen. "There are no papers for this, and I choose to disregard the countless number of people who say it is much easier for them to read if the line lengths are constrained as they are in a book or scroll or every other form of human writing ever put on this earth, so I will not make my site easier to read. F you."

I've a cousin who does something similar, and also (according to him) makes in the high hundred-thousands/low million range. In theory, it sounds like easy money, but I personally would probably break down in tears from boredom if I had to work on drop-shipping widgets, even if I was raking in the cash.