XuMiao

Most people have their own domains and don't want to make softwares. But their jobs depend on softwares sometimes. When the software developers don't understand the domain, it's reasonable to provide a DSL for users to customize by themselves. If the programming language is as natural as the human language, DSL is just the regular language.

After all, language is for the precise and concise communication, human to human or human to machine.

Most of the time, the surrogate metrics align with the unknown metrics when they are small, e.g. 80%. They can start to diverge at 90% for example. When I see people report a .5% improvement over a 98% accuracy and claim the statistical significance, I highly doubt about the real significance. The long tail is always longer than we thought. On the hand, there is no coincidence in the world. If some datapoints happen to be outliers that the metrics miscalculated, it might be that the problem isn't setup right. For example, missed some information in the data. In-person interview helps not just because people have a better judgement. It's because we can find out more details about that sample. The current AI/ML method lacks an active information seeking process. They don't have the freedom to explore the world beyond the defined state/reward space.

Exactly, the high dimensionality is very difficult to grasp for regular people. The nonlinearity of AI models makes the situation even worse.

OOP is essentially relational programming. It programs relations among objects. The encapsulation should be on the logic of relationship.

Transport is a relationship between passenger, viechle and state. Some passenger transports with a car, others by foot. It's the Transport's responsibility to update the states of passenger and viechle.

Encapsulation on entities sometimes is not enough to make a good OOD.

One key question is whether symbolic AI is the right model of the world. It underperforms vector based AI on many specific tasks. But human experts heavily reply on it to communicate with each other. If symbolic AI is not the right model, P vs NP problem might be just irrelevant. Human philosophy is full of craps. We will lose a lot of beliefs. Elon will be right, we will abandon human languages, and connect through a cable in the brain. Everyone will relearn every thing from NN.

If symbolic AI is the right model, but difficult to build algorithmically. Vector based model just help to make it faster and better. Then we humans are fine. We simply proxy the lower level optimization to AI. Our functionalities will be shifted just like what happened when engine was invented hundreds of years ago.

I am curious about that too. As a separate topic, if the operational dbs can be compatible with parquet type storage (backup and restore), the offline analytics and machine learning would be seamlessly integrated together. Offline analytics usually can simplify online analytics. Discovering new dimensions, normalization/denormalization, and optimization of indices and partitions. Operational dbs shouldn't have to stress themselves at the gunpoint.

Extreme Islamic is a difficult problem to deal with globally. Look at Israel, they are taking extreme measure too.

Historically, this extreme ideology comes from Western culture. Both Christian and Islamic are exclusive religions. Drives each other nuts for thousand years.

US played a very bad role in modern days and caused the situation out of control. Iraq war and other mediterranean wars, all of them are unethical anti-human-rights wars. They have destroyed so many families. That's why nobody in the world believe US's so called human rights propaganda even though there are some America truly care.

Chinese don't believe that US is able to solve their own extreme Islamic problem. They invented their own method. They believe that the fundamental reason for the extreme behavior is because they are excluded from the society and the economy. They want to reeducate them to get them involved. However, the approach is a bit harsh. Not quite sure whether they will succeed.

Frankly speaking, getting everyone involved in the society is the key for US as well. Rich people don't pay tax, don't participate in the economy, and don't care about poor guys. They have their robots. More and more poor people have left the economy completely for years. The middle class is vanishing quickly. The extreme behaviors are emerging fast these days.

Before we point our fingers at other people, we really should figure out our own issues.

CCP does have a lot of room to improve in protecting human rights. But to be fair, for the matter of national security, every country takes the extreme measure. Look at the Assange case. Neither Swedish nor US government makes any sense to me.

Unfortunately Capitalism is not Democracy. Capitalism serves a set of dictators instead of one dictator.

In fact, Taiwan has suffered from DPP for many years. The president Tsai has spent quite bit of funding to control the social media. They organize and spread rumors to attack other political figures. Double standard has become the new standard for the government. The laws, the judges, the elections, are controlled by DPP. People in the government, in the media do not care about the truth. They only care about the ideology, the party. Very typical results of the bipartisan politics.

It's going to be a turning point for Taiwan next year. They need a leader to pull the society out of this ugly bipartisan fighting.

Knowledge is a product, we should not only be a geek and leave the profit to the others voluntarily. All scientists should stand up to learn how to raise funds, market, sell, and refine. In fact, the business side of the effort is also very difficult and professional. The effort can further help reshape the research. Having a small team can reduce the bureaucratic requirements of prestige pseudo-scientific stamp but enforce the true scientific methology to solve the problems.

The business relationship among people is becoming flatter instead of deeper. Tooling is going to help individuals achieving more independently. At least it's the reality I would like to live in.

Hypocrisy is one of the originals of the Western culture. For example, I can be evil because the God will forgive me eventually. I can commit crimes and invade you but you can't because you need to be a Saint. I'm talking about the society as a whole instead of any individuals. This is the same as the Totalitarianism is one of the originals of the Eastern culture. These cultural traits date way before Democracy and Science, and start crawling back during the past decades. The Democracy was only invented several hundreds of years ago. It's an political ideal that's implemented with the Scientific methodology systematically. Without the sophisticated design over the system, the Democracy would never be successful. In fact, It's far from being mature and has been stalled for nearly half a century compared with the advancement of the Science. We need to put down our arrogance and emotional attachment to the good old days. We need to stop the Hypocrisy and reinvent the Democracy with the new Science. It sucks but it is necessary for the future of the entire human race.

Many aspects of this topic deserves extensive study. For example, ML is all about generalizability, ever since the deep learning flushed the field, it seems like the numeric representation (tensor) always yields better generalizability than the symbolic representation. Is it true? Or, under what circumstances, symbolic representation helps?

In the past couple of years, several papers have shown that predefined symbolic relationship can improve over vanilla DL. For example, recognizing a picture of numerical arithmetic equations and compute the result. This is very difficult for neural networks to parametrize over the pixel space.

Moreover, statistical generalizability is currently all derived from concentration theories. This means that the knowledge encoded by a neural network model depends on the statistical distribution of the data. If two people have two different sets of data, they might end with two very different models. Symbolic generalizability is quite different. A rigorous mathematical proof holds true as long as we all live in the same world with the same set of axioms. For example, no person can appear in two physical places at the same time. Or, if A causes B, A has to occur before B occurs. These knowledge can't be learned through statistical methods with no symbolic priors. We postulate the logic first, then verified through observations.

At last, problems that statistical learning handles well so far are essentially interpolations of the collected data. Being able to extrapolate well is still an unknown. Will inductive logic programing work better in this scenario?

Symbolic is the signature of human intelligence. All of our scientific breakthroughs are encoded by symbols, even the DL (deep learning) stuffs. It won't be replaced by the numerical paradigm completely in any time soon.

In fact, many quantatitive researchers in their domain are already data scientists without using modern tools like machine learning. Being able to use these tools is only the beginning of a new practicing and advancing.

I consider it the responsibility of computer scientists and engineers to make the tools better and easier. But unfortunately, it's that one has to become a good computer scientist first then a domain expert.

We are not doing enough on innovations of tooling.

Very reasonable comments.

I don't know why some people are so emotional about China. Sounds like they have lost something in their lifes.

An adult conversation is a fact driven conversation. That is the basic culture for HN v.s. places like reddit.

GDP is a manipulable number that has lost its accurate meaning. Even the ratings of the TVs have multi-aspects now.

I like the idea.

There are two types of ml jobs. One for modeling and another for computation. The former deals with the communication with human. The latter deals with the communication with machines. I would prefer that a statistical logic programing language (maybe extended from prolog) for the modeling part. A strong compiler tool or service that compiles it into any machines, on cloud or edge. Most of the time, what human likes to achieve is small and declarative. But for any small modifications of the program, we have to think about the computational flow, deal with the type errors and tune the low level efficiency of CPU and GPU. It's a waste.

I like the settings of SQL where modelers use SQL to mine data while engineers maintain the performance of the computational engines.

What do you think of life-long learning scenario that models are trained incrementally forever? For example, I train a model with 1000 examples, it sucks. The next guy pick it up and train a new one by putting a regularizer over mine. It might still suck. But after maybe 1000 people, the model begins to get significantly better. Now, I will pickup what I left and improve it by leveraging the current best. This continues forever. Imagine that this community is supported by a block chain. We won't be relying on big companies any more eventually.

AI will declare an independent government, all human will join that government eventually. We will be batteries in exchange of great life styles. No matter how hard we try, we can optimize the resources better than machines.

There is the mathematical logic that is deduced from axioms which is considered universally true unless the axioms are false.

Our everyday "folk" logic is always contextual and ambiguous. The deductions are probabilistic instead of being deterministic. They are sometimes true.

Americans killed native Indians and Mexicans. We made a lot of Chinese and African immigrants suffer for decades. The foreign policy is one of the sources of many troubles world wide including terrorism.

It's very easy to pointing fingers on others but self. Double standard is the weakest and most stupid culture characteristics from the West. No one will follow a leader like that.

I used to believe that generalization is about some nice concentration bound that one can design carefully in terms of the optimization algorithm. After many years practice, I began to firmly question about the basic assumption about data distributions. For many real applications, assuming the data are from the same unknown true distribution is fundamentally incomplete. In the real world, information is always partial and adversarials always exist. It never is a closed world where adding samples eventually has diminished value. Instead, it's an open world where long tail phenomenons are often observed. It never is about a single objective that the optimization algorithm cares about. Instead, it's entangled with many coupled but conflicting interests.

The existing statistical learning theory is having a hard time on addressing questions like combinatorial generalizability. Hopefully, the new stability analysis and adversarial analysis can lead us somewhere theoretically.

Without PhDs work, you have no clue where to look. Pushing the idea to market not only requires a market fit, but also requires BS everywhere to educate customers. Building a product are efforts, but it is as equal as the efforts of creating the idea.

In general, I agree that for avery complicated task that a machine intelligence performs far better than any human being, human-like reasoning might be just impossible. The logical formula might be too long to be reasonable to any human.

But from another side, current AI has a hard time to perform ordinary human reasoning at all. Those reasoning mostly contains a few logical clauses. Why?

Human like logic usually hide the common sense or context in other words. There is no absolute logic like FOL when we talk about human like logic. The traditional rule based AI system fails because of that.

On the other hand, deep learning are very strong at representing these hidden common sense and context. Look at the word2vec.

We now are at a point to combine these two approaches together not to build AlphaGo, but enable ordinary human like reasoning.

At least that is my hope.