kavalg

My kind persona: Cool! We all need this!

My cynical persona: Now sit and wait for the advertisers to start infesting HN with recommended videos :(

My pragmatic persona: Sounds like there may be some law related issues such as

- "Then we wrap it up in a YouTube UI" -> probably violating copyright. I can't help but remember that episode from Silicon Valley, where Dinesh dumped the company to Galvin to get rid of the COPA violation

- "I had to create some kind of API key from Google Cloud UI to query YouTube for video metadata" -> probably in violation of some ToS already

Good questions. I guess we will never know unless someone with enough money and data tries the approach on a larger scale.

Their premises are:

- If you work with the raw bytes in a non-hierarchical (isotropic) model, such as a byte transformer, you get better transfer to downstream tasks, but at a huge computational cost.

- if you use a tokenizer (pretrained or regexp based), you gain a lot of computational efficiency, but loose semantics and the transfer to downstream tasks may degrade

- finally, with H-Nets you get the best of both worlds. The model learns how to chunk the data in a better way than a hardcoded/separately trained tokenizer

Authors also demonstrate that this approach works in other modalities, such as DNA sequences and images, where we don't normally have lots of ready to use tokenizers.

50/50 is bad enough

Highly recommended! Sebastian does a great job explaining complex AI topics in a very intuitive way.

While I have the same sentiment, I am also curious why a competitor can't accelerate their developments with AI. In the end, hardware development is to a large extent also software development. Is it the patents that NVIDIA has that are preventing this? Or maybe some trade secrets? AFAIK the NVIDIA hardware is manufactured entirely by TSMC.

These are two different things. For example, it is possible that the fridges connect to a central, more secure MQTT endpoint to report telemetry and receive commands. They don't have to be discoverable on the internet. Also, not setting up a VPN, given the fact these vulns were known for so long is pure negligence.

yes, but why do they need to be exposed to the internet?

You guys are so funny :)

The most subtle thing to me is who has stake in the game. It is true that humans make mistakes too, but they would generally strive not to damage money-generating software/hardware infrastructure. If we remove all they people that actually understand how the system is working (up to a level of abstraction), how will the ambitious PO (or whoever is driving the AI) be safe from ruining the whole thing or simply landing in a deadlock situation where one agent wants to make the change (e.g. dev agent) and the other agent is refusing (QA agent)? I guess we are still trying to find the right level of abstraction and only evolution will show where it is.

This quote shall really resonate with all of you who use Tinder similar dating apps: "Commercial apps need you to keep swiping but this app is designed to get you off the platform. Singapore built the first dating app whose success metric is deleting its own users".

Even for a non-linguist like myself, the article didn't feel right. I expected to find more information about the Thracians, who inhabited those lands for quite some time and left some extraordinary (albeit non-written) heritage.

Since when has intelligence had a strong relationship to status and respectability? I've met intelligent people that don't get much respect or status either because they don't look good, are shy or they don't have money.

True, but the relationship is certainly not linear and not a Markov process. Also, the $$$ a company makes are rarely directly translated to incentives for developers. I remember a case where a company was doing an OK engineering/product job, but had problems with sales. Then a new sales driven CEO came into place and for the next 3-5 years the company had great financial results, while at the same time eroding the engineering culture (who cares if money is flowing anyway). Then that guy left and the next CEOs had to take care of the mess.

How come? the github page says the license is MIT.

I kind of agree with you at this point. When ChatGPT was rapidly gaining popularity I thought that they will eventually replace search (esp. for shopping), which would have given them a huge ad revenue. Maybe they could have even tried social networking e.g., to help you sort out the huge flow of information that today's social networks are and get to the important/rewarding/whatever posts. But now ChatGPT is kind of getting commoditized. I would even dare say that gemini feels to me a bit better now, so the search route for ChatGPT is clearly gone.

Yes, I can imaging that a robot chef holding a knife improperly is already dangerous enough even if it does not move (e.g., what if you trip and fall against the pointed knife)

Cool! Does anyone have a list of challenges that must be tackled before such humanoid robots become usable for tasks such as household chores, cooking etc?

It looks like the stance of FSF is for proliferation of the copyleft to trained LLMs

> "Therefore, we urge Anthropic and other LLM developers that train models using huge datasets downloaded from the Internet to provide these LLMs to their users in freedom"

In Azimov's Robot series the society that chose to live with robots gradually destructed itself by just living longer and not having so many children. The other part of humanity that avoided robots flourished (not without suffering). But that all required new planets for settlement (I am looking at you Elon).

MIT license, supported languages beyond english: ar, ch, de, es, fr, it, ja, pl, pt.

https://huggingface.co/HumeAI/tada-3b-ml

https://github.com/HumeAI/tada

Apprenticeship. You will have to prove to the company that working at a minimal wage is still beneficial. Or we can take it even further, you will have to pay the company for getting the necessary experience. Maybe you sign a 5 year contract with a big cancellation fee. It is not unheard of. I remember some of the navy schools having something like this. You study for 5 years for free (bed and food are paid by the school) and then you have to work for at least 5 years for the navy or pay a very big fine if you refuse to do so.

30+ minutes is a gross underestimation IMHO. It is probably somewhere in the range 1:10, 10:100, especially if you include the cost of context switches a senior has to do. In my experience, the loss of flow, due to context switch is very prominent and sometimes painful.

someone still will, but quality will become really expensive

Then there will be no one to buy the robots :)

I wonder why did they release it then.

Unless you can provide a (community) curated list of sources to search through (e.g. using MCP). Then I think local models may become really competitive.

You are completely right about the polyphonic expressivity of piano. What I lack is the intonation (bending) of tone.

This is very interesting! I can't wait to run it on some of my own compositions...

This is so cool and also MPL licensed! Thanks!