HN user
They are also pretty bad at navigating mazes (which can be somewhat similar in spirit to text adventures where you need to navigate through text): https://arxiv.org/abs/2507.20395
I stopped drinking coffee for four or five years. I drank a lot before. You do realise that it has a strong psychoactive effect, at least it did for me.
I still don't drink coffee, but I started experimenting with paraxanthine and I absolutely love it (paraxanthine is the primary metabolite of caffeine and is also a stimulant). I feel like it gives me most of the benefit of caffeine with very few downsides (no jitters, no crash, exits your system faster).
Oh the irony, I also felt that way when I started reading it. It borders on being hypocrisy given the topic of the article.
What are your favourite skills?
You can find a demo here: https://huggingface.co/spaces/facebook/omniasr-transcription...
Even if it was close to being near perfect, that is still not enough due to the negative impact of false positive detections on students.
That's where the humanizers come in. These are solutions that take LLM generated text and make it sound human written to avoid detection.
The principle of training them is quite simple. Take an LLM and reward it for revising text so that it doesn't get detected. Reinforcement learning takes care of the rest for you.
Why do the news need to make everything sound sensational. Let's move on.
The section on nested ifs reminded be of being a Nevernester: https://www.youtube.com/watch?v=CFRhGnuXG-4 (it's a short [~8 min], fun watch)
It makes it look like the presentation is rushed or made last minute. Really bad to see this as the first plot in the whole presentation. Also, I would have loved to see comparisons with Opus 4.1.
Edit: Opus 4.1 scores 74.5% (https://www.anthropic.com/news/claude-opus-4-1). This makes it sound like Anthropic released the upgrade to still be the leader on this important benchmark.
You would think that Springer did the due diligence here, but what is the value of a brand such as Springer if they let these AI slops through their cracks?
This is an opportunity for brands to sell verifiability, i.e., that the content they are selling has been properly vetted, which was obviously not the case here.
Taurine deficiency has been claimed to be a driver of aging [1]. The claim from the news article about it possibly being related to cancer seems like it needs a much stronger justification.
(watching live) I'm wondering how it performs on the METR benchmark (https://metr.org/blog/2025-03-19-measuring-ai-ability-to-com...).
It has happened to me on very large files and when you've had a long composer session. I usually check every file that is being modified for large removals because of this. It's a bit annoying, but something I can tolerate.
The original website is a news report of an article. The one he posted is from a peer-reviewed journal which has a much higher standard of reporting. The information there is reported by scientists with expertise in the field. You cannot expect the same level of rigour from journalists that try to sensationalise findings to get more clicks.
This post reminded me of Nevernesting: https://www.youtube.com/watch?v=CFRhGnuXG-4
Looking forward to see that project on air quality.
For what I follow, it looks genuine.
However, for many viral posts, many of the responses have nothing to do with the original post. It looks like bots/users that are riding off of the success of the viral post in an attempt to get more views/likes/followers. It's a sad behaviour that lessens the experience of the platform.
It has been grown in Iceland: https://www.nordicwasabi.com/
If you're using a Mac, I really recommend alt-tab https://alt-tab-macos.netlify.app/.
I love cmd+tab on a mac to switch between applications, but I was missing a feature to switch between windows of the same application. Alt+tab solves that in the same manner Windows does.
Thanks, that's great to know. Do you know how they would compare if you performed training instead of inference?
Isn't that going to be a problem with no NVIDIA GPUs?
Looks like mode collapse in Midjourney. It is less capable of dealing with instructions now likely to increase the chance of generating aesthetically pleasing output.
You may have an opinion, but can you justify why you don't believe chatGPT will achieve AGI?
Existing scaling laws show that perplexity (i.e., ability to predict next tokens in text) can be lowered by increasing model size and adding more data.
Why would a model that is better than us at predicting what words would occur next in arbitrary text not be an AGI?
Governments that implement these type of policies might see a rapid rise in their GPD. If it works out, I assume others would follow.
But what country would be the first one to take the leap?
I also recommend other books by Brian Christan, especially the Alignment Problem.
What made you decide to get tested for it?
Some people don't experience anxiety from caffeine whereas others do.
This effect has been linked to polymorphisms on the A2a receptor gene (https://pubmed.ncbi.nlm.nih.gov/12825092/).
Being a generalist is a lower risk strategy than being a specialist from a portfolio management perspective, i.e. investing how you spend your time.
If you specialize in something and it stops being relevant you might have little marketable skills to contribute. This would be like putting all the eggs in the same basket.
If you are a generalist you are protected from such a scenario since you put the eggs in several baskets.
But as a generalist you might be missing specialist opportunities so, of course, it's a bit of a tradeoff.
In this sense, it can be rational to be a generalist. However, every person's experience is unique so it's hard to say that one thing is better than the other without further context.
I have played really many boardgames but I think Isle of Skye is the game that is probably the one I would choose as my favorite one. It has a relatively simple auctioning mechanic and the point system makes it different every time.