We are still smarter than computers

This article is part of the On Tech newsletter. You can do this Register here Getting it on weekdays. We will leave for the rest of the week and return on Monday.

The idea of ​​computers bothering people has been a major focus of science fiction. My colleague Cad metz It is said that we are reaching there. Maybe.

Cad talked with me about a technology called GPT-3 that learns Crank out writing that looks like it was created by humans. This technology can help accelerate the development of computer programs that can spread conversations or throw information at the click of a button.

But Cade emphasized that technology has promised human-to-humans for more than half a century. It never gets there. or not yet.

Shira: Explain how GPT-3 works.

Cad: It is based on neural networks, a technical approach that learns by analyzing large amounts of data. This is how Siri recognizes spoken words and faces in Facebook digital photos.

In this case, a neural network is essentially scanned through the entire Internet: predict the next word in a sequence of words. But it also learned how humans connect language together. This means that with very little prompting, you can get it to generate a tweet, Pen romance column Or even write simple computer programs.

It sounds bonkers.

This is incredibly impressive, but the system has significant flaws. If I tried to generate a speech that sounded like President Trump, it would probably make no sense five times out of 10 of the GPT-3 version, repeating phrases in ways that people need editing to make logical Will not happen.

If something only works in half the time, it may not be useful. People go wrong in handling technologies as if they are actually more advanced because they are choosing successful experiments and ignoring mistakes.

What can GPT-3 be useful for, and an example of something potentially destructive?

An automated system that handles the repetitive, boring parts of writing software programs can actually make life easier for human programmers.

The most frightening possible use is disinfection. Imagine a computer that can create unlimited amounts of false information or fake photos and you cannot differentiate between real and fabricated.

We hear about technology all the time Writes like people or Talk like us, Or makes Realistic computer generated human face or Fake video. Why?

Since the 1950s, scientists have produced techniques that looked like we were close to imitating human intelligence. Every time we were nowhere.

The capability of such technology has advanced greatly in the last 10 years due to neural networks, but this human mimicry is largely based on detection and matching patterns. Humans do more than pattern matching. The question is, when will technology advance?

Well, are these things going to be as capable or more capable as people?

There is also a large argument about this among artificial intelligence researchers. Some of them believe that if you give neural networks enough time and data or computerized simulations of the world, they will eventually reach human intelligence. And others think it’s just ridiculous, at least in the foreseeable future.

So will you and I still write a job in 10 years?

After using this stuff and talking to a lot of people about this technique, I am not worried in the short term. But you can see the trend. This technology is constantly getting better, and there is no need to upgrade it further to supplement human workers or perhaps make certain areas obsolete for a longer period of time.

I would say that language models like GPT-3 are like self-driving cars. They both have improved dramatically but Last little correction Human capacity is very difficult to reach.

I am Caught on Tuesday about smartphone apps. But streaming apps for TV are even more terrible, as they mix the worst aspects of smartphone app stores and cable TV systems.

Here’s what I mean: If you have cable or satellite television service, every channel in your lineup is due to an agreement between the channel owner and your TV provider.

ESPN’s owner, Walt Disney, makes a complicated contract with Comcast and Comcast gives Disney a bunch of money, and then you get to watch college football games. Sometimes these business arrangements reach an impasse and ESPN may disappear for some time as Disney and Comcast fight over the money. This is annoying.

The streaming world could free us from the mess of this old school. Instead it messed up and moved it up a notch.

In the same way that Apple and Google decide which apps can work on your phone, companies like Amazon and Roku decide which apps you can download to your streaming TV device. And like cable TV companies – and unlike Apple and Google – Amazon, Roku and their partners often split airtime for in-app advertisements or pay streaming commissions for every Netflix subscriber who signed up for streaming services. Reach individual contracts with.

When lawyers can’t deal, you might not be able toO find hbo max, Peacock video app Or Apple TV + on your streaming device. This is bad That’s why you need Consult flow chart to watch tv, And streaming apps will probably cost more because they will absent this tangle.

I do not know what to do. Maybe we should all connect our smartphone or Our television computer And bypass streaming devices altogether. But know this in your heart: Streaming entertainment systems make no sense, and it’s not your fault.

  • More unwanted attention for YouTube: Google-owned site blocked right-wing Channel One America News Network from posting new videos for a week To break YouTube’s rules with false claims Regarding a guaranteed cure for Kovid-19. YouTube has specific policies to punish channels for spreading coronavirus-related misinformation, but not for false claims about electoral fraud, which OAN has aggressively pushed, my colleague Dai Wakbayashi writes.

  • We can free ourselves (slightly) from the screen: Set modest goals, such as a 20-minute daily limit for online news reading on weekends, and create a no-phone zone like a dinner table. My colleague Brian X Chen has these and others Suggest our brain to break through the wall in our phones and other screens. (Counterpoint to Brian: If I don’t look at the screen I think about the world and get sad and so give it to me.)

  • Use your precious screen time to stare smooth minds: Wired investigation memes that act as our attempts to laugh by screaming In a year full of disasters. “Unwritten Minds” Memes and Tiktok are videos that are astonishing “nee noo neo noo“Looks – to get out of thinking about what is happening around us.

Ping pong cat fell four stories From his owner’s apartment to the roof of another building. Ping Pong was rescued by London firefighters and he is well! Although you might not know by looking at it Ping pong’s expression in this picture. (Thanks to my colleague Erin McCann for sharing this on Twitter.)

we want to hear from you. Tell us what you think about this newspaper and what you want us to find out. You can reach us ontech@nytimes.com.

If you do not already receive this newsletter in your inbox, Please sign up here.

Source link

Leave a Reply

Your email address will not be published. Required fields are marked *