Chinese National Accused of Stealing AI Secrets from Google

A Chinese citizen who recently quit his job as a software engineer for Google in California has been charged with trying to transfer artificial intelligence technology to a Beijing-based company that paid him secretly, according to a federal indictment unsealed on Wednesday.

Prosecutors accused Linwei Ding, who was part of the team that designs and maintains Google’s vast A.I. supercomputer data system, of stealing information about the “architecture and functionality” of the system, and of pilfering software used to “orchestrate” supercomputers “at the cutting edge of machine learning and A.I. technology.”

Chinese National Accused of Stealing AI Secrets from Google

AI trade secrets are now the hot spy targets of many nations including China, Russia and others. It is truly not only the next battlefield but will ultimately determine who actually wins the "war" in my opinion.

In other words, "He with the best AI wins".

And what kind of "secrets" would that be?

AI is just a fancy new name for database centric programming, which most software engineers have been doing since the 70's.

The WWW has let it blossom into some very sophisticated applications, but down in the underpinnings are the same stuff we have been doing all along.

About ready for a new acronym, let's just call it BS.

a reply to: charlyv

Yeah AGI is still quite a way off in our future by my guess.

a reply to: SchrodingersRat

so you have to ask, did they get the woke one, or the normal one?

a reply to: charlyv

AI is just a fancy new name for database centric programming, which most software engineers have been doing since the 70's.

Maybe for expert systems, but database centric programming is nothing like machine learning or neural network systems. If you think they are, you are way behind the curve in your understanding of AI and I suggest your refresh your current knowledge.

ETA - You should really read the NY Times article I linked in my OP.

originally posted by: network dude
a reply to: SchrodingersRat

so you have to ask, did they get the woke one, or the normal one?

That's a good question. What was the name of that system, Bard or something?

a reply to: charlyv

Right? AI is such a bloated marketing term. It means almost nothing. Pong had AI. Madden '99 had "AI." Litterally every video game ever made has AI. It's the exact same concept. These models are not intelligently devising anything they weren't already programmed to.

a reply to: charlyv

I think AI gets overused and they slap it on things to make it fancy. A company wants you to use their chat bot instead of customer service so they say AI when reality is it’s no different than ten years ago with predetermined commands and functions.

But there are things like Chat GPT that are actually fed on large language models from all of the internet and it’s one interactions. There are tools using that along with audio and video that can make (almost) indistinguishable deep fakes.

What we’re seeing is it’s true infancy, and it’s easy to say it won’t go very far.

Windows 95 with the first widespread GUI was only 12 years before the iPhone.

A lot of doors are only now opening up due to hitting new computing capabilities and the widespread high speeds of internet.

I wouldn’t discount the possibility of high powered AI just because they’re going to use it as a buzzphrase. Just like I wouldn’t discount the ramifications of quantum computing if they truly crack that code.

Googles been pouring money into AI for a very long time, I would wager they have one much further along than they let the public see.

a reply to: CriticalStinker

And I bet they caused this double post too!

Those scheming AI's!!

originally posted by: Irishhaf
Googles been pouring money into AI for a very long time, I would wager they have one much further along than they let the public see.

They absolutely do.

DeepMind for starters.

a reply to: SchrodingersRat

I havent looked into it in years, realized it was a rabbit hole I just didnt want to go down.

a reply to: andy06shake

Well , the Chinese could maybe want to use that Stolen AI Tech in their Space Program but could run into some problems with it if they Stole the WOKE One . I can see it now ,....." Ropin the Pod Bay Doors Wong " .... Sorry Lee , I cannot do that , your Credit Score is Deficient . Please Try Again Later " ........

a reply to: Zanti Misfit

Or the new TikTok 2.0.

According to Moore's Law, processing power doubles every 2 years.

With A.I. that increase is every 6 to 9 months.

This creature that is being built is going to be smarter than us very soon.

youtu.be...

The genie is out of the bottle.

Elon Musk is going to court with Sam Altman and Microsoft in order to show the world what exactly is going on.

youtu.be...

originally posted by: SchrodingersRat
a reply to: charlyv

AI is just a fancy new name for database centric programming, which most software engineers have been doing since the 70's.

Maybe for expert systems, but database centric programming is nothing like machine learning or neural network systems. If you think they are, you are way behind the curve in your understanding of AI and I suggest your refresh your current knowledge.

ETA - You should really read the NY Times article I linked in my OP.

Please do not parrot that stuff on me, I recently retired as a senior systems architect and contracted to the military for many years.

Machine learning and neural network systems all have underpinnings of conventional, object oriented programming and the real data that they use is always supplied in source by humans. The abstraction is all based upon facts supplied by humans. There is no such thing as an "unsolicited interrupt", even at the hardware level because all exceptions are table driven by rules, rules put there by humans.

There is also no such thing as a machine having an unsolicited thought. We presently mimic AI, and yes, there are astounding revelations and sophisticated hardware and software, but it is not true "AI"

Some day, we will probably merge biology with silicon, but we are a long way from it.

a reply to: charlyv

Machine learning and neural network systems all have underpinnings of conventional, object oriented programming and the real data that they use is always supplied in source by humans. The abstraction is all based upon facts supplied by humans. There is no such thing as an "unsolicited interrupt", even at the hardware level because all exceptions are table driven by rules, rules put there by humans.

Of course all of the original info/data is put there by humans! Where else is it going to come from?

All AI is created by humans. It's what the systems can do after they've been given the data that matters. Especially non-linear, non-sequential operations.

I was simply commenting on your "database" line. Far too simplistic for these systems and this discussion.

And on a final note, I've been working with AI in many forms and in different capacities on and off since the late 1980's. I still do today, so you're not going to "out-tech" me.