Connect with us

Tech News

Oh God, DARPA is teaching AI to make more complex decisions

(TECHNOLOGY) All your fears about Artificial Intelligence are true. (Well, maybe.) DARPA is testing out common sense in AI, so hold your breath.

Published

on

darpa robot

There are two things we hear quite often. Someone in a staff meeting might make a proposal and defend it with, “It’s common sense.” Another’s rebuttal is usually the follow-up, “Common sense is not common.”

This has never been truer than my junior year of high school basketball when my teammate would yell “You’ve got a plastic bellybutton, Ref!” I’ll never forget looking at them in consternation to ask what it meant. They replied, dumbfounded “You don’t get it?” Let’s take this issue and scale it way up! The Defense Advanced Research Projects Agency (yeah DARPA, aka the government) is testing common sense in artificial intelligence.

Guys, the future is here. But how in the world can someone test something for machines that we barely know how to test in humans?

Artificial intelligence (AI) is great for simple, straightforward tasks i.e. my dad asking his phone “How old is Lady Gaga?” at the dinner table. It has now moved into anticipating i.e. responding to someone who emails that they are out sick with suggestions such as “Great!”

We can all see when we moved from the first example to the second, there was a slight disparity between the reliability of responses we would receive. That’s because logic and reasoning are things that can only be learned through a complex web of experiences. One only learns certain lessons after locking our keys in the car (definitely not while it’s running), placing a hot glass pan under cold water, and other embarrassing (and human) mistakes after we’ve made them.

So as terrifying and futuristic as it sounds, DARPA is looking to enable machines to make regular decisions, identify and analyze complex scenarios, and register solutions without the regular input of the human brain. While outcomes desired such as responding naturally to new situations and understanding the world around it sound fantastic, isn’t there a part in all of us thinking about every dystopian story where the machines rule the world?

Well, it seems we are a little ways off from that (for now). To begin, machines are given scenarios with multiple choices of what is likely to happen next. These are designed to prompt understanding and world-building so that fewer and fewer situations remain unrecognized. It has also been mentioned that this method is just the first phase of testing for common sense. The rest is to be seen.

For now, I supposed we’ll wait with baited breath and applaud the technological triumphs as they come. Let’s just make sure there’s a master safety switch.

Jennifer Yano holds a bachelor’s degree in Spanish and a master’s degree in teaching. A high school Spanish teacher and coach by day and a naturally reclusive Hobbit by night, she enjoys writing about business, education, and kitten paws.

Tech News

Degree holders are shifting tech hubs and affordability

(TECH NEWS) Tech hubs are shifting as degree holders move, but it’s causing some other issues and raising some interesting questions about the future of jobs.

Published

on

degree city

Bloomberg recently announced their annual “Brain” Indexes. The indexes are an annual reckoning of STEM (Science, Technology, Engineering and Mathematics) jobs and degree holders. The “Brain Concentration Index” approximates the number of people working full time in computer, engineering, and science jobs (including math and architecture.) It measures the median earnings for people in those jobs. It also counts how many people have a bachelor’s degree in a STEM field, or an advanced degree of any kind. It blends those things together to determine how “brainy” a city is.

Since they started in 2016, Boulder, CO has been at the top of the list. This year it’s followed by San Jose, CA, which many people might expect to be at the top. Many of the more surprising cities, like Ann Arbor, MI, Ithaca, NY, and even Lawrence, KS, are bolstered by the presence of a strong university.

It’s an interesting methodology. It’s worth noting that anyone with an advanced degree, whether it’s an MBA, a law degree, or a Ph.D. in literature, contributes to which city is a “tech hub.” It’s also worth noting how expensive many of these places are to live.

If you follow this kind of national data collection at all, you may also know that Boulder is one of the least-affordable cities in the country. So is the San Jose/Sunnyvale/Santa Clara metro area, with a median home price of 1.25 million dollars and a median household income of $117,474. (That means that the average mortgage is more than half of the average paycheck). However many people tech hubs like San Jose and San Francisco attract, they’re also hemorrhaging talent. Every day, 8 Californians move to Austin. Of the people who stay, more than half are thinking of moving.

They aren’t doing that for fun. As much flak as Californians get for gentrifying places like Austin, they’re being megagentrified out of their own homes. As salaries rise and CEO gigs attract the wealthy (and turn them into the Uberwealthy), the people who wait on tables or teach their children can’t afford to stay there anymore.

Speaking of people leaving, Bloomberg also measured what they call “brain drain,” the flow of advanced degree holders out of cities. They pair that with a decline in white-collar jobs and a decline in STEM pay to come up with their annual list. It includes places like Lebanon, PA and Kahului, HI.

All in all, it’s interesting information. But there are other factors at work that it can’t speak to. What does wage stagnation in the U.S. mean for the flow of education workers? If San Jose and San Francisco can be tech hubs based on the number of people with degrees, but people are still fleeing, what does that say about rankings like these? What human stories get lost in the shuffle? And is “tech hub” even something a city wants to be if that means running out of teachers (or making them sleep in garages)? Where does the next generation of tech hub workers come from?

Knowing the people behind the numbers makes it clear just what a mixed bag this is. Maybe we need more tech hubs like Lawrence, Kansas. Or maybe we need rent control. Or maybe we need to embrace remote work. Maybe there are no answers. As interesting as data like this is, there’s something sort of wistful about it, too.

Continue Reading

Tech News

New Apple Watch is awesome, but past watches could be just as good for cheaper

(TECH NEWS) The Apple Watch Series 6 is a ridiculous display of self-flattery—but that doesn’t mean people won’t line up to buy it in droves.

Published

on

Apple Watch being worn on wrist showing weather for Montreal.

The Apple Watch has been the subject of everything from speculation to ridicule during its relatively short tenure on this planet. While most have nothing but praise for the most recent iteration, that praise comes at a cost: The Apple Watch’s ghost of Christmas past.

Or, to put it more literally, the fact that the Apple Watch’s prior version and accompanying variations are too good—and, at this point, too comparatively cheap—to warrant buying the most recent (and expensive) option.

Sure, the Apple Watch Series 6 has a bevy of health features—a sensor that can take an ECG and a blood oxygen test, to name a couple—but the Series 5 has almost everything else that makes the Apple Watch Series 6 “notable.” According to Gear Patrol, even the Series 4 is comparable if you don’t mind forgoing the option to have the Apple Watch’s screen on all of the time.

More pressingly, Gear Patrol points out, is the availability of discount options from Apple. The Apple Watch Series 3 and Apple Watch SE are, at this point, budget options that still do the job for smart watch enthusiasts.

Not to mention any Apple Watch can run updates can utilize Apple’s Fitness Plus subscription—another selling point that, despite its lucrative potential, doesn’t justify buying a $400 watch when a cheaper option is present.

It’s worth noting that Apple is no stranger to outdoing themselves retroactively. Every year, Apple’s “new” MacBook, iPhone, and iPad models are subjected to extensive benchmarking by every tech goatee around. And the conclusion is usually that buying a generation or two behind is fine—and, from a financial perspective, smart.

And yet, as the holidays roll around or the initial drop date of a new product arrives, Apple invariably goes through inventory like a tabby cat through unattended butter.

The Apple Watch is already a parody of itself, yet its immense popularity and subtle innovation has promoted it through several generations and a few spin-off iterations. And that’s not even including the massive Apple-specific watch band market that appears to have popped up as a result.

Say what you will about the Series 6; when the chips are on the table, my money’s on the consumers making the same decisions they always make.

Continue Reading

Tech News

Microsoft acquires powerful AI language processor GPT-3, to what end?

(TECH NEWS) This powerful AI language processor sounds surprisingly human, and Microsoft has acquired rights to the code. How much should we worry?

Published

on

Code on screen, powering AI technology

The newly-released GPT-3 is the most insane language model in the NLP (natural language processor) field of machine learning. Developed by OpenAI, GPT-3 can generate strikingly human-like text for a vast range of purposes like bots and advertising, to poetry and creative writing.

While GPT-3 is accessible to everyone, OpenAI has expressed concerns over using this AI tech for insidious purposes. For this reason, Microsoft’s new exclusive license on the GPT-3 language model may be a tad worrisome.

First of all, for those unfamiliar with the NPL field, software engineer, and Youtuber, Aaron Jack, provides a detailed overview of GPT-3’s capabilities and why everyone should be paying attention.

Microsoft’s deal with OpenAI should come as little surprise since OpenAI uses the Azure cloud platform to access enough information to train their models.

Microsoft chief technology officer Kevin Scott announced the deal on the company blog this week: “We see this as an incredible opportunity to expand our Azure-powered AI platform in a way that democratizes AI technology, enables new products, services and experiences, and increases the positive impact of AI at Scale,” said Scott.

“Our mission at Microsoft is to empower every person and every organization on the planet to achieve more, so we want to make sure that this AI platform is available to everyone – researchers, entrepreneurs, hobbyists, businesses – to empower their ambitions to create something new and interesting.”

OpenAI has assured that Microsoft’s exclusive license does not affect the general public’s access to the GPT-3 model. The difference is Microsoft will be able to use the source code to combine with their products.

While OpenAI needs Azure to train these models, handing over the source code to another party is, to put it mildly, tricky. With the earlier GPT-2 model, OpenAI initially refused publishing the research out of fear it could be used to generate fake news and propaganda.

Though the company found there was no evidence to suggest the GPT-2 was utilized this way and later released the information, handing the key of the exponentially more powerful iteration to one company will undoubtedly hold ramifications in the tech world.

Continue Reading

Our Great Partners

The
American Genius
news neatly in your inbox

Subscribe to our mailing list for news sent straight to your email inbox.

Emerging Stories

Get The American Genius
neatly in your inbox

Subscribe to get business and tech updates, breaking stories, and more!