Crossover: 2014

Chapter 180 The opposite opponent

Face recognition in previous lives took off after the emergence of artificial intelligence.

Of course, just artificial intelligence is not enough.

Face recognition technology involves the intersection of artificial intelligence and big data, and may also involve things like cloud computing.

All in all, this thing is not very suitable for Lin Hui in a short time.

If you look at it in the short to medium term, Lin Hui should still be an artificial intelligence.

There has been a lot of thinking about artificial intelligence Lin Hui before.

Although the direction of artificial intelligence may not be the best entry point, it is the most suitable for Lin Hui.

After all, Lin Hui of artificial intelligence is quite familiar.

In his previous life, Lin Hui often dealt with things involving artificial intelligence.

Moreover, Lin Hui has already made a lot of preparations for the direction of artificial intelligence in this time and space.

To some extent, Lin Hui's layout in generative summarization algorithm paves the way for officially entering the field of artificial intelligence.

The reason for this entry is that the generative summary algorithm is an interdisciplinary subject in the field of natural language processing and machine learning.

And this branch of discipline happens to have a low threshold and a very low lower limit.

It's normal to intervene in this way.

But this sub-discipline has a high upper limit at the same time, and it is also very good to use it as a direction to break the game.

Although Lin Hui can break the situation with the help of the subdivision of generative summarization algorithm.

But the breakthrough achieved here is only an academic one.

The academic breakthrough means that Lin Hui can grab his academic status to a certain extent.

But this is far from enough.

A mere academic breakthrough is not enough to fully mobilize the enthusiasm of the market.

Although for ordinary scholars, an academic breakthrough can already be very exciting.

However, from the first patent published by Lin Hui, Lin Hui was destined not to be a pure scholar.

When it comes to artificial intelligence, Lin Hui is interested in Qian Jing of artificial intelligence.

Who would not be tempted to change a trillion-level market?
It is conservative to say that the trillion level is conservative. Lin Hui remembers seeing such a description in a work report:

"The scale of the core industry of artificial intelligence is in the trillions, and the accompanying economic benefits exceed ten trillions."

It can be said that Qianjing and prospects are very broad?

Even if there are some deviations between the two time and space, this kind of big strategy can only be about the same.

However, if you want to stir up the enthusiasm for the commercialization of artificial intelligence, just academic status is not enough.

In the final analysis, academic status can strengthen the appeal to talents.

But in order to maximize commercial benefits, it is also necessary to find the real explosive point of artificial intelligence.

Like a bunch of theoretical research by Einstein, ordinary people have no way of knowing the details.

But as long as Einstein's research is related to "traveling through time and space".

Ordinary people may still not understand Einstein's theory very well.

But this does not prevent people from understanding Einstein's theory.

In this way, "traveling through time and space" is the breaking point of Einstein's theory.

What is a burst point?
The hot spot refers to the point that can ignite the emotions of the people who eat melons.

The emergence of explosive points can make the general public interested in technology.

Even if it doesn't interest the general public.

At the very least, the head of capital/governor must be interested.

If the funder is not interested, how can he happily collect money?
It is good to use generative summarization algorithm as the entry point/entry point in this aspect of artificial intelligence.

As long as the logic is self-consistent, academic closed-loop growth can be completed smoothly and smoothly.

But if you start from the point of making the funder interested.

The generative summary algorithm is destined not to become the explosive point of artificial intelligence.

It seems impossible for a group of funders to see the broad money scene of artificial intelligence from the aspect of generative summary algorithm.

In fact, when it comes to the generative summary algorithm, let alone let the funder know what it is.

Many practitioners are also confused.

Not long ago, Lin Hui saw the journal of Princeton University saying:
The Department of Mathematics of Princeton University and Google SEARCH formed a research team on a new summary algorithm.

The research team claims to come up with a new algorithm that is more efficient than Lin Hui's algorithm.

Project title of this team:
——"Efficient summarization algorithm based on LSTM long short-term memory neural network"?
Seeing this name, Lin Hui was a little dumbfounded.

Two very top teams obviously want to get a better summary algorithm, why do they have to work hard on the long-term short-term memory neural network?

Lin Hui was a little speechless for a moment, and the only adjective that came to mind was "contradictory".

In Lin Hui's impression, the "long short-term memory neural network neural network" was originally proposed to deal with the gradient disappearance and gradient explosion problems that may be encountered when training traditional cyclic neural networks (however, this problem has not been solved)

Although the long short-term memory neural network may have certain advantages when dealing with long sequences of text.

But it's obviously difficult to handle the generative summary algorithm just by relying on this thing.

Taking this as the research direction, Lin Hui doubts whether these people can achieve any results.

And this time, Google Research is cooperating with the Department of Mathematics of Princeton University.

Lin Hui did not miss some academic forums in his previous life.

Although I didn't learn much serious knowledge.

But there are a few anecdotes about Princeton's mathematics department that I know a thing or two about.

The mathematics department at Princeton University is full of eccentric and stubborn people.

Lin Hui feels that if the Google Research Institute cooperates with the Department of Mathematics of Princeton University, there is a high probability that it will be led astray.

What's more, the research direction of the two is still in the direction of LSTM.

Lin Hui felt that it was a high probability event for the two to conduct cooperative research and deviate from the course.

But Lin Hui couldn't relax either.

Appropriate overestimation of opponents is necessary.

A person's success should be based on a stronger foundation.

Instead of counting on the opponent to be better.

But there is one thing to say, these followers are pushing in the wrong direction, which also shows a reason.

That is to say, the direction of generative summary algorithm is really not suitable as the explosive point of artificial intelligence.

Things that are easily confused by professionals.

But a project that is so easy to be confused like a generative summary algorithm is obviously not suitable as a breakthrough point for artificial intelligence.

Could it be explained to the general public a little bit?

In the fast-paced era, not everyone is so patient.

What exactly is suitable as the explosive point of artificial intelligence?

and many more.

How did artificial intelligence become popular in the previous life?
Lin Hui suddenly thought about this question.

Thinking of his previous life, Lin Hui suddenly had an answer:
——AlphaGo AlphaGo.

Although the research on artificial intelligence in the previous life has a long history.

But let's talk about when artificial intelligence exploded.

It seems to be something after 2016.

In March 2016, AlphaGo and the Go world champion and professional nine-dan Go player Lee Sedol had a Go man-machine battle and won with a total score of 3 to 4.

After that, artificial intelligence suddenly became famous.

Can something about Go become the explosive point of artificial intelligence?

(End of this chapter)