tornikeo

> The volume of a sphere grows faster than the surface area. But if traversing the interior is instant and frictionless, what does that imply?

It's nearly frictionless, not frictionless because someone has to use the output (or at least verify it works). Also, why do you think the "shape" of the knowledge is spherical? I don't assume to know the shape but whatever it is, it has to be a fractal-like, branching, repeating pattern.

Meaning a GPT but next token is a live sensor reading or a servo angle or accelerometer state. Then connect that GPT with an actual LLM as a controller and you (hopefully) have a physical machine with arms, legs and a mind.

I think it's a combination of multiple factors. I worked with GPU kernel codes before and the code that you write has a tendency of never being updated or modified. once it works it works perfectly and you do not change it. if you get new hardware you're going to fully rewrite it. so, typically readability is just not useful. also, you're never working with variables that make sense to humans. it's never something tangible. it's always tiles, offsets, indices. i do not think, at least when I was writing the code for GPUS to waste space visual space on better variable naming was worthwhile.

Not related to this article, but this post seems to attract so many cranks. Just look at other comments. The amount of weird (out-of-touch) comments on this post is fascinating.

I love this. I think of mathematics as writing programs but for brains. Not all programs are useful and to use AI for writing less useful programs would generally save humans our limited time. Maybe someday AI will help make even more impactful discoveries?

Exciting times!

I did not expect that ending.

From "I've spent the last six months working on a deep learning system to improve virtual screening for drug discovery"

To "I’m currently working on a new startup in the blockchain space with a couple co-founders."

I don't get why invest time in PhD if your work afterwards seems totally unrelated to your expertise. Is this how most of the PhD stories end, working at a completely-unrelated-to-your-expertise job for a good pay?

Typst is almost as easy as markdown, but my only gripe is with the web editor app: the inability to insert links quickly, i.e. with a copy-paste shortcut.

Use the best tool for the job.

I look at LinkedIn as a purpose-built tool for marketing my brand (my dev experience). It works well. Just like any tool LinkedIn has some problems. But it is the best in that niche.

I'm thankful for the US administration to truly give other countries a chance to come forward and something good, and take the center stage.

But, as good as that opportunity is, I'm afraid the 3 years this administration has left wouldn't be enough to fully disassemble the US advantage. :)

What happens if I upload an image of me and my (real life) partner? Would this app add some third wheel? :)

> First-author paper published at a top conference

I now find that requirement in most AI-related high-tech jobs. Starter salary for these jobs is often $150k+. When someone is willing to pay you $150k+ for having published a paper, fraud definitely makes (financial) sense. Basically, the problem is the demand, and the demand corrupts the metrics (h-index).

I think that having a paper published would become less and less significant in the future. With time, businesses will also move onto other signals of success.

>People deserve to live with dignity, earning a living wage.

Look, I get where that is coming from. But no one deserves anything, even clean air. Instead, you pay for things with other things (e.g. clean air in exhchange for slightly more expensive cars).

A (shitty) job is better than no job. Minimum wage should be lowered. You gotta help people with something else. For instance by allowing to build higher density housing.

Garbage in, garbage out. If the training dataset (accidentally) paired silence (`X_train`) with `رجمة نانسي قنقر` tokens (`y_pred`), then any silence will always be translated to that. Fortunately, this particular problem is easy to fix--just detect and remove silent parts before API call. This also has a side benefit of saving you money on transcription.

I think we are having a Deep Blue vs. Kasparov moment in Competitive Math right now. This is a large progress from just a few years ago and yet I think we still are really far away from even a semi-respectable AI mathematician. What an exciting time to be alive!

This reminds me of how I used to have spam-filled email inbox before I switched over to GMail. It almost feels like we are back to that state. There's now a large demand in keeping the context of humans free of AI bullshit. I wonder what the solution to this would look like? Identity-based blacklisting?

Unfortunately, i don't have the cold hard data on what job complexity distribution looks like right now. But I don't think it's far fetched to assume that most days, most tasks people do are route and boring. Possibly a fraction of those tasks are amenable to AI-base acceleration. Who knows? But we do know (from that article) that people are hesitant to even try using AI.

It seems that the average employee doesn't currently benefit from completing their job more efficiently; i.e., they have no interest in AI.

To accelerate mass AI adoption in workplaces, it may be necessary to expose the average worker to the risks and rewards of business ownership. However, it might be the case that the average worker simply doesn't want that risk.

If there's no risk, there can't be a reward. So I can't see a way AI adoption could be sped up.

No, it is not time to fork HN into AI/LLM and "Everything else/other".

I enjoy the website as-is, and simply use search when I want to get to the topics that interest me.

Straight from FAQ [1]:

> Who can see my data? > Only you can see your data. Almost all of it stays on your computer and only a little bit goes to our safe helper. We also give you ways to make it even more private if you want.

So, am I correct to understand that the actual model backing this is on the server side?

[1] https://meeting-buddy.com/Faq

I'm sorry, could you clarify how "Fine" is different from copilot/Cursor or even ChatGPT?