If you think any model coming out of china is not containing as much stolen data as humanly possible from the actual innovators u crazy.
Using a chinese model is like watching pirated films. If you actually support the people who make the thing you love then you don't do that.
We then get hit with chinese sponsored propoganda of "the big labs stole artist images to begin with so they can't complain" and act like LLMs didn't create an entire new category of thing.
So much propaganda so many lies and dishonesty because at the end of the day China knows it's not innovative enough to compete. It hopes to hobble our abilities while they frantically steal to catch up.
But the truth is that if China ever gets a lead over US in any innovation they will instantly kill any openess. If chinese AI ever passes US it's likely you'll never touch that model ever.
Just like the russians they use our own culture against us and it's absurd people fall for it.
A foreign government state sponsoring IP theft and manipulating markets for unfair competition.
Here's a 1:1 example if the USA did exactly the same thing.
1. CIA and Ford work together to steal mercedes IP
2. Gov gives money to Ford to build the same car half the price
3. Throw as many roadblocks at mercedes regulation wise as possible until it is not financially viable to sell in US anymore.
4. Sell the same car back to germany at half the cost they have it because mercedes isn't state sponsored
5. Crush mercedes who created the better car to begin with.
That's not only dishonest but actually harmful for innovation and economic growth. China has done this Sooooo many times it's absurd.
They don't have nearly enough compute to be competitive on pretraining and distillation is exponentially cheaper.
It's literally the same tactic they do in every industry. Steal top IP, gov funding, ban the company they stole from domestically while driving cost to zero, try to kill the original IP creators.
They are actively in the process of doing this with Tesla as we speak. And are in the phase of pushing them out domestically and trying to destroy tesla marketshare globally.
They have obvious skills and add a lot of value while doing this but US could stop all of this by simply stopping their supply.
If anything the leadership gave BS answers to the journalist for the exact reaction HN is portraying right now.
"They're so dumb AI is harmless" is a danerous take when the people in charge are often times more educated than an average westener. They also are decent at propoganda (you have certainly consumed pumped or direct propoganda from terrorist orgs).
Their leaders are often university grads from the west. It is smart (enough) people leading delusional farmers but the article is clear that the farmers are just given AI commands from the leadership. There's still strong asymetry occuring here that more funded orgs have the better AI but if AI flattens and opensource catches up it's gonna be a real interesting world where every terrorist also has a team of advanced weapon engineers and tacticians available.
I did large scale tests temp 0 and there was still randomness with the same prompt inputs coming in.
I did this with several model apis.
GPU processing is not going to be the same from what I read but also the AI backend is doing a lot of fancy batching resulting in another layer of randomness.
But we have regulation and complaince for consumer secrets? That's not a comparable example.
The difference is that if openai gave you a product and it leaked a million peoples bank passwords it would destroy the entire company.
Again until a big tech product can bring that to a clean user experience we're not there yet. Even the most zealot openclaw users are not hooking their bank accounts into the AI yet. I'm sure they exist but I've not seen them.
Also every big tech computer use product actively screams for you not to give their agents secrets.
The iphone moment is an AI that can completely manage your personal life. It has full access to every financial account you own handles all admin work. Could sign you up for a new account pay and give you the login.
If you can SAFELY do that it's a big moment. But to be clear safe is a massive problem. Until you see a big company start saying the AI can use your SSN, CC, bank password safely we aren't there yet.
I am saying it's one of many examples of which I personally experienced and those I know have also experienced.
And yes it should and does. If an industry does not provide value it naturally fails because it doesn't make money. Academia has no profit by design and now gets to face its own form of regulation for disfunction.
There is no way to even attempt to fix our science base without first cutting much of it. How are we supposed to fix a system while not applying consequences to it? Every US institution is rotting from inside out demanding more money while providing less. The most effective way to begin the process for any of these is to cut money and sort out real value.
This is like US public transport argument. Should we just forever fund disfunctional systems because we are scared of trying to fix them? Our public transport is beyond disfunctional compared to basically any other country on earth.
Since the market cannot correct these systems as what makes industry so effective then you must artificially apply equivalent market forces.
If the leader of one of the countries most famous research groups commited research fraud with a slap on the wrist it's not hard to realize how broken it is.
I have personally worked with famous researchers and seen a few fudge their work.
Acting ignorant of the major flaws implies you haven't been involved, or inside a bubble of research nirvana.
Using a chinese model is like watching pirated films. If you actually support the people who make the thing you love then you don't do that.
We then get hit with chinese sponsored propoganda of "the big labs stole artist images to begin with so they can't complain" and act like LLMs didn't create an entire new category of thing.
So much propaganda so many lies and dishonesty because at the end of the day China knows it's not innovative enough to compete. It hopes to hobble our abilities while they frantically steal to catch up.
But the truth is that if China ever gets a lead over US in any innovation they will instantly kill any openess. If chinese AI ever passes US it's likely you'll never touch that model ever.
Just like the russians they use our own culture against us and it's absurd people fall for it.