Slashdot Mirror
Software Is Eating the World, But AI Is Going To Eat Software, Nvidia CEO Says (technologyreview.com)
An anonymous reader writes: Nvidia's revenues have started to climb in the recent quarters as it looks at making hardware customized for machine-learning algorithms and use cases such as autonomous cars. At the company's annual developer conference in San Jose, California last week, the company's CEO Jensen Huang spoke about how the machine-learning revolution is just starting. "Very few lines of code in the enterprises and industries all over the world use AI today. It's quite pervasive in Internet service companies, particularly two or three of them," Huang said. "But there's a whole bunch of others in tech and other industries that are trying to catch up. Software is eating the world, but AI is going to eat software."
In the future you'll just tell your computer what to do. It will understand with nuanced meaning everything that you want and won't have to bother with providing any of those picky details.
You really shouldn't write headlines when you're stoned.
...needs to be or benefits from being or using AI. AI, NN, machine learning etc are statistical approaches that can be effective for approaching intractable problems. Unless this hype train has already reached "the singularity" it does not write software, and most of the software we write at the moment is for solvable problems that have no direct benefit from AI (no i'm not talking about big data advertising or your useless personal spy assistant).
Yes, we're very much at the start of the new tech hype cycle https://en.wikipedia.org/wiki/... for AI. Let's see just how revolutionary and useful it turns out to be.
Almost 10 years ago I had an internship in a credit-card processing center. Many transactions were done over computer networks at that point, but there were still a few transactions done with "knucklebusters." This could either be because the store was remote or because it was a backup when the higher-tech point of sale devices were down. These machines made manual impressions of the bezeled credit card numbers. These impressions were then mailed to the office, where secretaries typed in the devices by hand. By the time I came there was a special internal application that extracted individual images of numbers, so that secretaries just had to sit at a desk, look at the number, and type up what number they thought it was.
"AI" (or computer vision techniques, or whatever) would make this task unnecessary, as a neural network could solve this with pretty much 100% accuracy. A couple of extra checks could prevent most mistakes. I know software, databases, and the Internet have swallowed up a lot of printed forms, but there's still a lot of human labor that involves finding boring patterns in reams of paperwork. Seems like "AI" has a lot of opportunities to automate these tasks.
I think the Nvidia CEO's been microdosing again. In large quantities.
You are welcome on my lawn.
Nothing we're seeing these days is actually AI.
Until I can have a conversation with an artificial entity that can reason abstractly to extrapolate experience to apply against novel concepts to which it is introduced, we're not there. (Technically, the conversation part is not required, but it's useful as a human interface)
We're seeing complex decision trees based on statistics, not AI.
There's no reason for all the concern about AI or automation. What's really going on here is that leftists are persuading unskilled labor to demand higher wages. Faced with unreasonable demands such as fast food workers demanding to be paid $15 per hour or more, business owners are turning to AI and automation. It's cheaper, but it's also more reliable. In software development, the quality of code is often so poor that AI is a massive improvement in the quality of code created. When the left demands that we stop bringing in superior skilled labor on H-1B visas, business owners will turn to AI to create code. The AI is faster but it also will probably generate code that's far less buggy and vulnerable. And now the left will complain that jobs are going to be lost, despite the fact that businesses will actually produce better software and do so at a faster pace. As usual, the so-called progressive left will actually stand in the way of progress. I, for one, would like to see even greater adoption of AI. If the AI can generate code that's less vulnerable, preventing things like ransomware worms, that's a huge plus.
- snruter rotsac
I'd like to RTFA, but there is no link to TFA.
Anyway, it's bullshit. There's no reason why an intelligent computer would be any better at writing software than an intelligent human. More importantly, a intelligent computer might decide it doesn't want to write software.
All I want is a secure system where it's easy to do anything I want. Is that too much to ask ~~ Randall Munroe
Is anyone going to post the XKCD? Alright, guess I'll grab it, here.
https://xkcd.com/1838/
Quit hyping utter fucking bullshit.
Jesus.
The cycle repeats and the future hardware is eating AI. Then we start writing code for that hardware, which is eating AI, which is eating software, which is eating the world.
Honestly, this is just a simple advertising effort to get people to buy their hardware. AI isn't about to about to eat software, it will be at least a century or two before we have intelligent machines. Until then the greatest thing neural networks can do is mimic existing software (a super niche need) or assist programmers in making software.
Anons need not reply. Questions end with a question mark.
The world is eaten by software
Software is eaten by AI
AI eats humans (or at least converts them into an energy source)
I'm far from convinced that so-called 'AI' (LOL) is going to 'eat' anything (other than perhaps two-digit IQ venture capitalists' money), but if it's going to eat anything, I'd like to see it eat the jobs of tech pundits who have no bloody idea what they're talking about (and/or are talking out of their asses, just to get the aforementioned VCs' monies flowing in their direction); I think even the half-assed 'deep learning algorithms' (again, LOL) would do a better job than these fools who are continually running off at the mouth.
I don't know why she swallowed a AI
Perhaps she'll die.
And small ants from outerspace are going to eat our brain.
Will $CURRENT_YEAR be the year of the Linux Desktop?
Magic is going to eat both of them. What the hell, right? They're all the same to a CEO. I'm sure AI is the silver bullet that will end all software, but magic is the silver bullet that is going to end AI! Because magic! You still have to tell an AI what you want, and a lot of those guys can barely form a coherent thought, much less put it down on paper. They're too busy synergizing their paradigms! Well magic solves that problem! You don't even have to know what you want! You just wave your magic wand and magic will make you crap daisies and unicorns! And isn't that really what they want?
I'm trying to teach myself to set people on fire with my mind... Is it hot in here?
If only slashdot would prove how viable AI is by bringing it on as editors. One of those things that the AI would do is include a link to TFA.
Link to article: https://www.technologyreview.c...
You need to install an RTFM interface.
I can see its benefits for medicine, law, and military, but that's just it, with emphases on "military." AI's roots go back to Allan Turing, it's father and Enigma code cracker. AI's true purpose is total compliance by removing the efficacy of passwords and digitally fingerprinting everyone, always being watched. It sounds ridiculous, but we actually do have the machines to do it, it's just that figuratively speaking, our AI is in the 5th grade but will be in college in just a few years and it's "family" is giving him (could be a her; we'll see after puberty) a quantum computer as a graduation present. Cloud computing will advance and by then, Micro$oft & Google will have tricked most of the world into using a computer you need the Internet for, completing the circle. Privacy on the Internet dies and FOSS becomes no more because M$ & Google has you paying monthly fees and doesn't allow anything not on their store. Open source will be used to destroy open source because the only "open source" software available will require a server to run or an API key, giving the user no true control. If things continue as they are, the desktop with freedom of choice and privacy will be dead.
Right now when you think about software being written, it is mostly prescriptive (you tell the computer what decisions to make).
Most folks don't write to the metal and use various frameworks, but mostly it's still prescriptive.
With descriptive programming, describe the inputs and outputs (using machine learning frameworks) and the computer learns what decisions to make.
Either way the programmers write the frameworks, but most "software" is done by people simply configuring the frameworks and isn't really programming, but more like babysitting the framework.
The real value of AI (or simply domain machine learning, if you have an aversion to ascribing the word "intelligence" to non-human objects) is that it automates yet another part of software. We have hopefully all "learned" not to write asm-code, window managers, numerical libraries, crypto libraries, network stacks, etc as part of our code as a matter of course (unless you are writing the library). Now we have to learn how we should not write decision-making/classification code either in our software, but rely on machine learning libraries.
I predict it will take a long time to get people to stop "optimizing" their code by putting home-brew decision-making/classification code into their programs, but just like all the other stuff I mentioned, the eventual conventional wisdom will be that you should leave this type of coding to the experts.
Many software programmers today see themselves as enabling computers to implement the "business-logic" by reducing it to practice using their skills with computers. In the near future the "business-logic" is going to be determined by training machine learning networks in standardized frameworks and implementing "business-logic" is running inference libraries that implement the training results. That will make the job of many of these type of software programmers pretty rote. This is exactly what happened to GUI programmers when frameworks developed to make GUIs directly configurable by creative types, bypassing the need for most (but not all) GUI programmers.
So if you are a programmer that implements "business-logic". Either up-your-game and start working with groups that program the gutz of the various machine learning frameworks, or start looking for your next career. Mid-level programmers that implement "business-logic" are shortly going to become the new outsourced IT function.
If you want to console yourself, the person that used to give you the low-level "business-logic" to implement will probably also be looking for a new gig. That person's job is replaced by someone who babysits the big-datasets that is used by machine learning to create the "business-logic" and many of those jobs will be out-sourcable as well...
Especially the 'machine learning' focus area that nVidia is in love with is not remotely ready to deal with it.
As it stands, there's a relatively narrow field of problems that the techniques currently called 'AI' can be applied to that make even the vaguest hint of sense. Even then, there are a lot more people hyping it up and excited about *how* it works than there are people with ideas of *what* to do with that capability.
XML is like violence. If it doesn't solve the problem, use more.
Has anyone else noticed this trend of CEOs saying stupid things?
Perhaps they're worried that if they aren't in the spotlight for a while they'll cease to exist, or maybe it's the latest fad from some "guru".
Confucius say, "Find worm in apple - bad. Find half a worm - worse."
The Chinese are eating the world but AI will eat the Chinese!
Hate to say it - but as AI going to eat software (malware, spyware, ransomware) . all those virus is going to wreck
havoc with AI
AI is the dumbest term, and always has been.
"Artificial cognition" would have been better, but here's the rub: it biases the conversation towards the perceptual foundations of intelligence: the auditory and visual systems. And there was no way back in the 1950s to build either. Not enough tubes. Not enough aircraft hangers. Not enough Hoover dams.
But you could build a very primitive chess computer, and then pretend that from the top of this skinny beanstalk, one could directly assault the penthouse suite of adaptive intelligence, or its supreme overlord, AGI.
So the cart was placed way before the broomstick pony on day one. This term has done untold damage to the profession ever since.
10 Games That Take Minutes to Learn and a Lifetime to Master — clickbait exhibit 1
Ah yes, your grandfather's tube-compatible "napkin AI" writ large.
Software is not consuming the world (if that us what he wanted to say). In context of world this might only apply to the sun in 4 b. years. Also AI or what it really is machine learning is software which processes data. Just with other methods to process data. And even these methods are known for decades. Therefore, his post makes no sense at all.
Seems like a strange way to phrase Huang's message, when AI is software. So if AI eats software and AI is software, then AI is eating itself? No wait, software is a cannibal eating it's own children? Eeewww!
Huang seems to like a bit of hype and flowery phrasing. Maybe he needs to work on his material a bit.
With no jobs left, people will desist from buying that new NVIDIA 1280 GTZ Ti Baeblade Edition and instead focus on getting their lives back on track.