swframe2

I saw an interesting quote from the deekseek CEO:

"When tackling a task, experienced people might reflexively tell you 'this is how it should be done,' while those without experience will experiment repeatedly, think deeply about the best approach, and ultimately find a solution tailored to the actual circumstances."

I do not think that OpenAI is out of ideas, but I do believe that large businesses have difficulty overcoming the innovator's dilemma. In addition to big businesses, there are also many small and very smart companies that are using investors' over exuberance to their advantage. Nevertheless, there are enough people trying that some are likely to find a way to develop higher-quality models at a lower cost.

trump spreads his income perfectly across all his LLCs. Their expenses always exactly match their income to the penny. I wonder why he is able to do that?

Please note that the 24H2 update will remove the Mixed Reality Portal and Windows Mixed Reality support. If you have a VR headset it might not work after the update. There are articles that claim that steamvr is still supported but it didn’t work for me.

When you run their demo locally, there are two places that trigger a warning that the code loads the weights unsafely. To learn more about this issue, search "pytorch model load safety issues" on Google.

I tried using Hunyuan3D-2 on a 4090 GPU. The Windows install encountered build errors, but it worked better on WSL Ubuntu. I first tried it with CUDA 11.3 but got a build error. Switching to CUDA 12.4 worked better. I ran it with their demo image but it reported that the mesh was too big. I removed the mesh size check and it ran fine on the 4090. It is a bit slow on my i9 14k with 128G of memory.

(I previously tried the stability 3d models: https://stability.ai/stable-3d and this seems similar in quality and speed)

Just accept that we all are part nature and part nurture. I suspect your good friends accept you as you are; some probably really like you as you are. It is also kind of relative; Hang out with the some people and you will always feel stupid or smart. I think you should focus on doing things that you and others find very interesting. There is a famous line from Dune that basically says "fear is the mind-killer"; generalize that to "thinking too negatively about yourself can lead to poor outcomes".

I clicked on a bad link a few months ago. I can't believe I fell for it. I've disabled javascript by default in my browser and only enable it for websites that I trust. It is painful for some websites that redirect a lot.

What are you doing to reduce your chances of running bad javascript code?

Related maybe, I got a new phone number and I was able to use it to sign into the previous owner's account at a site that uses 2FA. I noticed it when I tried to setup a new account for myself. The website wouldn't allow it but it happily allowed to me access the previous owner's account. It was inconsistent, removing the phone from their account required their password but changing many other important fields didn't.

To build an LLM that can reason about the 3d world. I suspect they will add the ability to reason about the physics of the world next. It's just another attempt to get closer to AGI.

They most likely will have to pivot a few times but once they show their LLM solving problems that others can't, the others will quickly add these features too. Right now, it is cheaper to wait for World Labs to go first. The others are not that far behind: https://cat-4d.github.io/

We were told that nothing can escape a black hole.

But when black holes merge, the combined mass is smaller.

It is described that the lost mass is converted to gravitational waves.

But gravity is the curvature of space caused by mass, it is not described as a form of mass.

How do you understand this process of matter escaping a black hole?

Is it that the gravitational waves are caused by a "quantum gravity" particle that can't be converted back to any of the other quantum particles?

MS removed support for mixed reality in windows 11 24H2 which made all windows VR headsets unusable (except MS's headset). I wonder if that counts.

I saw a recent video about the errors of alphafold 3.

https://www.youtube.com/watch?v=E61wJXlENoE

I think more generally, the issue is that people want to believe lies.

The Veritasium youtube channel has a few interesting videos related to this topic:

This one is on IQ: https://www.youtube.com/watch?v=FkKPsLxgpuY Pay attention to the stats on how IQ correlates to success (near the end).

This one is about becoming an expert: https://www.youtube.com/watch?v=5eW6Eagr9XA

This one is about someone who just worked harder than everyone else: https://www.youtube.com/watch?v=AF8d72mA41M

Is there data on how the parts move and deform due to human actions such as walking?

It is suspected that "fiction spreads faster than facts" on social media. It is also suspected that "when you present the facts to someone who wants to be believe a lie, that person will hold on more strongly to their belief in the lie".

It might be easier to use chatGPT to build something that is worth $10M. For example, there is the guy that made a crypto-currency.

(a crazy idea I wanted to get feedback on) Wouldn't it be much easier if everyone licensed tech from space-x? A space tech franchise haha? (To build a colony on mars we need a massive number of flights, why not spread the costs across several nations? Why spend $2b to $6b per flight on a space-x competitor; we want redundancy but it seems massively cheaper to license rather than reinvent. )

This youtube channel intuitively explains algorithms on a whiteboard: https://youtu.be/IqXdjdOgXPM

(I don't have a background (or interest) in quantum computing) I've tried some of the "chat with pdf" websites for other topics. It seems like they could help you understand the paper. If you give it a try, could you report back here on how it worked for you.

I think voters deserve the government they elect.

At times, it feels like we're cruising for a bruising but natural selection seems to be the only solution we can agree on.

I am curious. Why not give each app a private copy of common user resources? Every app has access to contacts but by default only the ones they create. Then android should allow sharing across apps based what the user wants to share. It would be a little bit tedious to share but an OS provided sharing tool can reduce that friction.

(slightly off-topic) The elephant in the room is ... Don't we need to address the issue that openai uses the data for training their future models? For many inputs this is not a big problem at all. But shouldn't there be an option (even if it requires an additional cost) that allows users to enter a mode where their input is not retained?

(my opinion) It is not predicting based on 'words/tokens'. It is transforming the general words/tokens embeddings into a context specific embedding which encodes "meaning". It is not an n-gram model of words. It is more like an n-gram model of "meaning". It doesn't encode all the "meanings" that humans are able to but with addition labelled data it should get closer. I think gpt is a component which can be combined to create AGI. Adding the API so it can use tools and allowing it to self-reflect seem like it will get closer to AGI quickly. I think allowing to read/write state will make it conscious. Creating the additional labels it needs will take time but it can do that on its own (similar to alpha-go self-play).

It is related to subpixel labelling. When a line/curve in the foreground is smaller than a pixel you end up having to edit the mask one pixel at a time. The authors of DIS are working on a new dataset and model which should work for my use case.

BTW, I used DIS to create the labels of a batch of 20 images, I manually corrected the labels and used them to fine tune a new model. That worked well but still it took me several hours to edit labels.

I tried using stable diffusion generated labels several weeks ago but I think with controlnet and other advances I should try again.

(My dataset is about 100k images. I probably only need to label about 10k to fine tune DIS).

The best solution I've seen is https://github.com/xuebinqin/DIS. You should try the DIS example images at the SAM site.

The main issue I have with DIS is that creating the labels of my own dataset is super expensive (I think it might be easier to generate the training data using stable diffusion rather than human labelling)

I saw the news earlier and only just realized we worked together. He was an amazing person.

History has shown that selfish behavior should be expected. The dangers seem unavoidable.

I think this application of gpt is way more useful than the chat interface. (Add an option for users to pay sooner). Here are a few suggestions:

1) Allow it to take a search term, do a web search and allow the user to select from those results.

2) Allow it to look at more than one document.

3) Detect if the output contains math formulas/graphs and render them. (or allow me to write a javascript post processor so I can add that logic myself)

4) When a user question can't be answered, prompt the user to allow your system to web search and then include those documents.

5) Create a version that can be run locally for those of us with private data. You should charge a lot for that version (~$100+k if the customer provides the hardware, and $1m+ if you have to provide the hardware (blackbox)).

6) Detect research papers and read the citations. You may have to ask the user for a SSO key to get the citations from paywalled sites.

7) Abstract responses need to be made more concrete. See if you can train the model to provide an example or describe the purpose or intuition when it responds.

With AI tech getting dramatically better quickly, we need to wait wrt restructuring the dominant players.

Basically, all the major players in the tech space are going to make major transformations to their internal services over the next 5 years. YouTube as you know it today is not going to be here 5 years from now.