_mitterpach

You might have a problem with this approach - the waste is large, the environmental impacts terrible. But they got to where they are exactly by this process. See articles such as SpaceX Launched Over 80% of All Orbital Payload Mass in Q1 2023 (https://www.nextbigfuture.com/2023/05/spacex-launched-over-8...)

Europe could stand to learn a thing or two. We still do our launches in secret, often not telling a public that there will be a launch until the satellites are already in space.

What a line. Reddit is well known of its position, and I am often one of those 100. Will this cease to be if they push further monetisation? Probably, if a good enough competitor, with good enough of a culture can stand up. Until then, not one other social network can replicate this effect.

Reddit killed forums, perhaps forums could kill Reddit now?

Honestly, fair to him, he caught the interviewers with their pants down, made an absolutely amazing promotion for himself and created a useful product that points out a terrible societal error. That's some hacker spirit.

I'd honestly be interested in his MRR now, it has to be at least on the level of an internship wage, right?

I can't say that I have not used it in similar ways though, and it might be very often the best alternative. I just worry that in this age of misinformation, many incorrect 'facts' will be absorbed blindly into these models and spewed forth.

(edit - wrong source, my bad) [1] https://nypost.com/2025/02/28/business/sam-altmans-openai-la...

What will be the reasoning taken, what are the guidelines for AI to make this choice? If, as a government official, you are using AI for any action, you should be required to publish the model you are using and the specific prompt you are using.

An analysis could then be run on many different seeds for the model and prompt - some will label you as a 100% traitor to democracy, some will label you as a fighter for justice. That is almost certain.

Now what is the extent you are comforable with? Should we ban somebody whose actions are marked as traitorous in 10% of seeds? 20%? 80%? Do you actually believe anybody in the government, making these decisions, sees this nuance?

I don't believe this is a right move for Google, but Google does not care about that. Google cares about showing ads, and I'd love to know how they plan on doing that, when everybody will google a term, read the (terrible) AI description and close the tab. Will the ads be injected directly into the responses? Will this be fairly and transparently disclosed?

There was a running theory that google results have detoriated in quality because google wants you to scroll down, to look at more ads. This runs contrary to this expected behavior, so I don't know what to think. I might have to start looking for an alternative.

I really like Ecosia, but if anybody has any recommendations, do tell.

I recently watched the 'documentary' about Gabi Petito's murder on Netflix. At three hour-long episodes, it was a well made series, with lots of background information. Even after these three hours, I could not really comprehend everything that happened, and I could not possibly make an opinion on this tragic event.

After she died, the documentary showed the media response - instant, with lots of talking heads shouting their opinions, many of which have been debunked very well by the documentary. But the commentators did not necessarily care about facts, nor nuance, they just shouted their opinions violently and outragedly.

Context is everything, nuance is everything. This is a big loss for current media, as everything has to be cut down into the smallest possible size, just to fit on Reels, TikTok, Shorts and other short-form content. If it is not in the form of a witty tweet, it is worthless and has no value.

I really thing there is no such thing as black and white, that there is not one thing better than other. There is no singular topic on which you would find 100% agreement. There is no topic on which you would find 70% agreement There is always somebody who will have some other angle on a topic.

I don't know if the current form of journalism is better or worse. I'd guess a bit worse, but there is nuance. You can find information (and misinformation) a lot quicker. You know about more that has happened in the world. Is that good?

20$ for basic access, OpenAI is losing money on it. 200$ for the most advanced access, OpenAI is losing money on it.

20000$ for specialised AI agents, which are not actually going to work as an agent and will need to be sheparded by a senior developer. I beg you employers, just hire juniors. Invest in human assets, invest in people. This industry is doomed otherwise.

This is not necessarily unique for Britain and feels like more of a world-wide trend. Of course, this can be partly explained by women reading a lot of romance books, I think it could be a worrying trend for the future. Reading is an important aspect of a man's life. Sitting down and being forced to uninterruptedly focus on a singular task is exactly the kind of thing that is missing from many lives of young men.

All of humanity's problems stem from man's inability to sit quietly in a room alone. Blaise Pascal

Simply put: - Police Enforcment unfairly policed black neighborhoods, overrepresenting them in historical records - Machine learning models were run on said historical records - Police Enforcement used said models to justify their policing, saying it was algorithmically determined.

I may be blanking on a few details, so if anybody can add anything please do, it has been a while since I've read it. These laundering algorithms will cause untold damage and should've been banned immediately.

Balatro is an absolute indie darling and one of the greatest games of this decade. I still think it not getting the GOTY was a miss, this should be a wake-up moment for the industry. Gameplay talks. Not micro transactions, not ultra realistic graphic, gameplay. If your game is fun, people will play it.

Hope many more games like this will come out in the upcoming years, if the industry learned anything.

Could perhaps someone, preferrably from an HR space, confirm or deny this presumption of mine?

There is many an open source software that could actually use a little bit of copying from the thing they are trying to emulate. GIMP is still one of the ugliest programs I have installed on my computer.

The design is very similar to Discord, could this possibly even go to a copyright breach or is the bar for that set too high?

Current process is very simple and direct, and you can go from seeing a game on Steam to installing it in under a minute. That is a huge win for customers and publishers both. I do wish there was still a way to replicate it, and I do laud the few game developers who do put their manuals on Steam, as the service definitely offers it.

According to Microsoft, users can ask questions about their PC, such as "How do I set up a Bluetooth headset on this device?" and Copilot will respond with something tailored to their device rather than a generic web search.

If only our systems could be designed in a way that every choice user makes is obvious and easy to find. Or make the search menu actually work. Dealing with Microsoft products is like dealing with an adversary, who is really keen on making my life more difficult at every turn.

And sadly it seems that their focus is still on ending support for Windows 10, to which I now feel a sort of Stockholm Syndrome.

Boom says its analysis has found that no audible sonic boom reached the ground during the flight.

Minimizing sonic boom has been a key goal for engineers involved in the race to bring about the return of commercial supersonic air travel.

That's interesting. Could anybody explain how they actually managed to do that? I always thought sonic booms were just this quirk of physics that is more-or-less unchangable, is their method of supressing such booms just flying higher, or is there some technology advancement that allows for this?

Building on top of this would be a terrible idea. The bubble is going to burst, OpenAI will stop getting investment money and will need to actually start being, you know, profitable. They will do that by milking all their actual existing customers - expect price to at least double in the following years.

I don't believe they have paywalled any core features, only paywalled a few new features that are cosmetic only. If you don't pay for Discord Nitro or anything, you can still chat, call, join and use servers. The biggest advantage of Nitro is actually shared with everyone on your favorite server, and by 'boosting' the server you give features to everyone on it. I like that.

Is there any way that they could have monetised it better? I see it as pretty creative and user-friendly, given there's not many avenues to monetise a free service such as a chat.

I think the new changes just bring it more in line with the rest of the market, and the features it does have, like continuing ad-blocking, are just too nice to pass up.

I do like the new vertical tabs introduced in the latest version, see Firefox 136 Desktop gets Vertical Tabs (https://news.ycombinator.com/item?id=43254871)

According to Google translation, “scanning electron microscopy” in Persian is “mikroskop elektroni robeshi”, while “vegetative electron microscopy” is “mikroskop elektroni royashi”. They are only differed by a point in the Persian script:

میکروسکوپ الکترونی روبشی vs. میکروسکوپ الکترونی رویشی

I see the difference and it is laughably small, but it is not a typo when the problem stems from incorrect translation, which was not caught by any review, which has now mad its way to two dozen scientific papers. At that point it should be looked at what else has made its way past the verification process, that such nonsense can end up printed.

On the other hand, I guess a part of scientific process is a public verification, so by catching it we're doing our part in science?

OpenAI and every other peddler keeps promising upgrades in the magnitude of 100-200x. That's impossible. To be affordable and usable enough, they would have to come up with actual miracles.

At this point, they've picked off all the low hanging fruit. At best, we're going to be seeing marginal improvements at the cost of a large price increase (see latest OpenAI flagship).

There is some usefulness, but after so many dollars wasted, all companies are going to be left with is a bunch of useless GPUs. I can't wait.

Chain-of-thought is well understood to be the way to squeeze out performance out of these models. Slow down, go step by step, use more tokens to get a little bit better output. This is so useful, that often in the comparision graphs you see between different models, the best one (the one I want to sell) is using chain of thought while others are not. Not always, but companies have been caught using this technique to hype up their models.

We’ve already seen that CoT (usually!) improves a model’s performance. So does prompting the model with examples of correct question-answer pairs (called in-context examples). But in reports, some models are evaluated with CoT, while others aren’t. The number of in-context examples is often different, and the prompts are almost always different.

https://asteriskmag.com/issues/07/can-you-trust-an-ai-press-...

As soon as OpenAI was desparate for performance upgrade, they implemented this as an actual feature.

However, a quirk in the way Unicode handles emoji has meant that some of the symbols take up many more characters than others. For example, a flag can take up as many as 14 spaces in your Tweet. Twitter has announced that it's changing the way it counts emoji so that they're all counted equally, as two characters. - https://www.theverge.com/2018/10/11/17963230/twitter-emoji-c...

So 4 character flag, which is written as 1 character, is sometimes counted as 2 characters. Honestly, I don't know, I did not expect for this to go so deep.

The article is stating interesting things, but does not really provide any alternatives. Current sites are located in places so as not to pose any danger to human settlements, and rockets are flown mostly over sea or desolate deserts. Sure, more could be done to protect the wildlife, but there are standards that all launchers have to abide by. If such standards are too lax, that is for regulatory bodies to decide.

Some estimates suggest the number of launches per year could hit 1,000 by 2050, up from 112 in 2020.

Deciding policy based on the guess that current number of launches will 10x in 25 years may not be.. wise.

If somebody was strangled by a shoelace, should we look at how the shoelace was imported?

Focus on mental health, focus on defusing situations, focus on better policing and protection. This is useless.

First thoughts, I have to say I like it. I'll probably find something wrong with it, but I've always found it an interesting layout, I might stick to it for the time being.

I think there is a large market for apps that just work, they do one thing and one thing well. I remember reading about somebody who makes a living out of their app, which is just a crossword maker, but accessible enough to be used by teachers and ordinary people.