Fully open reproduction of DeepSeek R1, which ends marketing hypes of USA big techs

You over estimate this AI variant and its impact on current exploitation efforts.

DeepSeek will not unseat the leaders and it will take some time before it is understood well enough to actually do anything useful.

While I think that DeepSeek will have an impact, it's going to be on par with other advancements in the field that already were great on their own (for example: Llama), I agree with the sentiment that it's not going to be as dramatic as people are making it to be. It's absolutely astounding work, for sure, but it's not "David defeating Goliath". Sadly every new thing is reported as an Earth-shattering event with world-changing consequences that will absolutely change life as we know it. That's always a lie, and that's what gets the panic going and people (and markets) overreacting.

And talking about blowing things out of proportion, I would also love if companies would start being more realistic with how they present the models. They are good as text manipulation tools that seem to hold some amount of knowledge and concept abstractions. They are good at extracting core ideas, commands or actions from natural language, executing on them, and then reporting back in somewhat-natural language, so they are great for personal assistants and text-processing aids. But that's not how they "sell" them. They advertise them as if they were a cosmic all-knowing thing that will "boost" productivity and reduce the workload of people everywhere, and that's absolutely not even near to being a reality.

Now, back to DeepSeek. For models to be profitable and actually useful without becoming a major disaster both financially and environmentally, they need to start using less resources, and that's exactly where DeepSeek presents a move in the right direction. The fact that people can download it from a repo (thanks for sharing one that seems to streamline some of the process) and run it without requiring too expensive hardware, will hopefully mean that more and more people will be able to play with this kind of tool and maybe find a better use for it than replacing customer support on their products and then finding out in court that it wasn't a great idea.

As a personal assistant collecting web information, these AI models can replace certain human force.

For instance, if you hire a secretary and ask him/her to collect certain information on open web, these AI models could perform better.  Other than these, AI service cannot replace an human secretary.

As to creative art, AI model can do very little.

Sure but there's nothing particularly intelligent about it.  That's just collecting and collating information under some given rules.  I expect they might perform better only in the sense of (a) doing that task way more quickly than a human; and (b) being able to access and filter more data in a given time period than a human.  The best I'll say about it is that the algorithm underpinning it is clever and the natural language processing has become useful because computing power is significantly better than it was 40+ years ago when NLP started being investigated.

Eliza started to be written in '64. That's 60+ years ago!

It was, and obviously very basic given the time period.  It's crazy to think that people used to think they were speaking with a real human and even formed emotional attachments with it.  I find it amusing that it was written in a programming language called MAD-SLIP!  I think AI really came into its own as a serious research topic in the 1980s although there were initiatives before that.  Shows how far we've come in all that time.  Not!

I got into it in the 70s, and programs like SHRLDU were already considered "old hat"...  but it really started to break out (in terms of hype, anyway) in the early 80s. But considering serious AI lies in the intersection of psychology, systems, linguistics, brain studies, philosophy, ... it really hasn't been that slow. Of course, two AI-winters (so far) hasn't helped - a side effect of the perpetual hype cycle.

HKPhysicist said:

As to creative art, AI model can do very little.

I used AI to create this image for an e14 contest:

Microsoft is being sued by some creators as well.  It even steals some actress's voice.  

As to myself, I never publish any ground-breaking scientific ideas in open web.  The recent progress in language model has proved that my worry is realistic.  OpenAI simply claims that my ideas are their inventions.  They can silence me whenever they want.  

HKPhysicist said:

Microsoft is being sued by some creators as well. 

Because it uses office 365 user content to train its model?

So far what I see is people are proud of their AI initiatives and will readily brag about how much it improves their productivity, even if it is just used to generate generic responses such as thank you notes. One thing to worry about with most on-line AI is that you are giving as much info as you are getting.

I use ChatGPT daily, but I mainly use it to get quick answers.  I try and resist getting it to do 'all of the work.'  You make a good point when you say, "you are giving as much info as you are getting."

I use ChatGPT daily, but I mainly use it to get quick answers.  I try and resist getting it to do 'all of the work.'  You make a good point when you say, "you are giving as much info as you are getting."