Paul Chen

Founder of WinningHealth AI lab
Researcher in AI in healthcare, HIT, Biomedical engineering, etc.

[Paul Chen @paulcx]

@pmddomingos Yes, the data statistics of a basketball video were completely wiped out.

[Paul Chen @paulcx]

@vivilinsv @JeffDean Do you have a full video?

[Paul Chen @paulcx]

@ClementDelangue From another perspective, Apple might be correct about doing nothing.

[Paul Chen @paulcx]

@Akashi203 @runinfrai The large model is open source, and deploying it costs $60. Seriously?

[Paul Chen @paulcx]

@bleysg @NVIDIARTXSpark Does it support Docker deployment?

[Paul Chen @paulcx]

@rbhar90 @AnthropicAI Agreed. So I don't understand why these top scientists and engineers would believe that a signature is effective. Or are they just expressing an attitude?

[Paul Chen @paulcx]

@kimmonismus Unless top labs like OpenAI, Anthropic, and Google disclose their most advanced technologies, how can developing labs trust them?

[Paul Chen @paulcx]

@ChrisGPT I think this is similar to many international agreements. Developing labs won't trust it unless Anthropic first opens-source the strongest technology, proving that this isn't a delaying tactic.

[Paul Chen @paulcx]

@CatChen 你可以用一个空头的公司搭建一个免费的路由平台，用这个公司控股一系列子公司，并保证每一个子公司的收入不超过$20m？

[Paul Chen @paulcx]

@KengGuangLong 一个简单的逻辑问题，为什么垂类模型不能是基于通用Agent的微调。除非垂类任务已经简单到通用模型也可以完成，那么不仅仅是垂类模型，整个垂类商业模式也将不复存在。

[Paul Chen @paulcx]

@natolambert Agreed. I seemed to see Robin Hood's story come back again.

[Paul Chen @paulcx]

@doodlestein If one day they take over your source efforts and use administrative power as tools against you, would you care?

[Paul Chen @paulcx]

@samhogan There's a Chinese idiom called "a target for all arrows," and now it's called anthropic.😂

[Paul Chen @paulcx]

@julien_c It's okay to engage in light banter, but their attitude and value behind are annoying.

[Paul Chen @paulcx]

Andrew's comments pointed out the problem. This guy is so called a "pig teammate" in the Chinese context.

[Andrew Ng @AndrewYNg]

@Mononofu @JensenHuang This is a false equivalence. Everyone has the right to keep their code private. The problem is when someone tries to stop OTHERS from open sourcing.

[Paul Chen @paulcx]

@ai_for_success I suggest also training this part into the open-source model, so that all AI can learn who their enemies are.

[Paul Chen @paulcx]

@Hyperion__13 @BingLiu34173809 单位距离问题不是吗？

[Paul Chen @paulcx]

@AlexKontorovich It was actually very tense, for example, Wang Hong eventually forgot to shake hands with the last person.😂

[Paul Chen @paulcx]

@alexframegreen @SecScottBessent Students attend classes as distillers, and we watch TV as distillers too. A technical term is used for commercial attacks—it's quite dirty.

[Paul Chen @paulcx]

@stephenandstuff @jenzhuscott There are many complex reasons, such as experts may not understand the annotation of specific tasks, such as evaluating or scoring certain processes. Each expert also needs training. This kind of industry is not mature, etc.