What are bots?
An internet bot is a software application that automatically runs over the web. Usually, bots are performing the most simple and monotonous tasks on the speed many times higher than it would be possible for any human. The variety of tasks bots can do is fascinating: starting from inoffensive Instagram bots aiming to help people with their account growth and ending with the huge networks of “droids” used for DDoS attacks. How many bots are operating on the web? Much more you might think. Here are some shocking data: according to the latest survey, more than a half of all website visitors are bots: after examining of 16.7+ billion visits to 100,000 domains, which were selected randomly, only 48.2% of captured visits were made by a human. Moreover, the vast majority of traffic is made by bots.
How long do they exist?
Creating an automat which is able to imitate human cognitive agency is one of the oldest dreams of humanity. Several authors claim that roots of this idea can be found in Ancient Greek mythology and literature: such famous figures as Hephaestus and Talos have some features of...basically, bots. The first widely-spread ideas of mocking natural intellect’s activity could be found among Modern European philosophers, such as Rene Descartes, Gottfried Leibniz and Francis Bacon. Nowadays the main concept is common: even Instagram bots resemble the basic idea.
The 1950s: Turing Test
The discourse of Artificial Intelligence (AI) was spread in the XXth century, after the range of innovations in the mathematical logic and informational technologies. Such famous mathematicians as Alan Turing and Bertrand Russell were concerned with this problem. When it comes to Turing, you have probably heard of his famous test. It was developed in 1950 in order to designate an automat’s ability to exhibit intelligent behavior indistinguishable from a human one. The idea is as simple as pie: a human is addressing the same questions to a machine and another person. His/her task is to determine which answers are automatically-generated. If he/she fails, the automat has passed the Turing Test.
This criterion - an ability to imitate natural language conversations - was questioned by latest cognitivist. They came up with the differentiation of so-called Weak and Strong Artificial Intelligence. The first one (also known as “narrow intelligence”) is focused on a precise task and only able to plausibly imitate human intellect. Bots are an iconic example of it. But when it comes to Strong AI (also referred to as Artificial General Intelligence (AGI)), it should be capable of experiencing consciousness and applying it to any kind of problem. We don’t have any examples of AGI yet: it is still one of the biggest concerns of contemporary science.
The 1960s: first chatbots and pessimism
The next significant step in the history of AI was the emergence of ELIZA -- a bot, whose purpose was to mock a psychotherapist. It was developed in 1966 by MIT professor Joseph Weizenbaum. The technology was war far from perfect due to ELIZA’s limited vocabulary. You can still test it here.
So, it is not a surprise that there were lots of skepticism about machine learning. What is more, 1966 ALPAC report was so rife with pessimism about the perspectives of AI research, that it affected government funding in this sphere. It was a crucial and unpleasing moment for the industry: there were only a few favourable developments until the 1970s
The 1970s: Artificial vision and androids in medicine
The first attempts to apply bots in such a sophisticated sphere as medicine were made in the 1970s. MYCIN was developed in 1972 by Standford Medical School. Their creation was able to ask a bunch of the most basic questions a doctor would ask to complete diagnosis. What is more, it was capable of identifying infectious diseases -- with the help of database made by experts. That was followed by the emergence of INTERNIST-1 -- comparable development made by University of Pittsburgh. Unfortunately, the elaborations were ceased since then.
Another worth mentioning episode of the history of bots was connected with Freddy -- a non-verbal robot, which was so interesting because of his capacity to “see”. With the help of the camera, Freddy was able to recognize different objects and operate on them. The main drawback was its speed: it took almost a day for Freddy to complete the most simple tasks.
The 1980s-1990s: the breakthrough into everyday life
With the increasing popularity of the internet, bots are finally left research centers and appeared on average internet user’s daily routine. The history of web bots starts back to 1988 with the uprise of Internet Relay Chat (IRC). The early IRC bots were primitive and used for imitating activity in games’ channels to save them from shutting down due to inactivity. In the middle 1990s, bots started to be used as web crawlers for the early search engines. One of the pioneers was WebCrawler, developed in 1994. What is more, the famous Googlebot originated from BackRub, developed back in 1996.
The 2000s: the Golden Era of chatbots.
The year 2000 was cardinal for chatbots and their communication with people. That’s the year of SmartChild emergence. This bot was programmed to search for answers to any kind of questions. We can call it the forefather of Siri -- of course, the contemporary voice search tools are much more sophisticated.
The range of tasks bots was able to solve was rapidly expanding. One of the noticeable examples is IBM’s Watson. This chatbot was specially designed to participate in the quiz show Jeopardy! in 2006. The machine finally succeeded in 2011 by winning the first place prize.
What are the most common ways to use bots around the Internet?
So, what bots are doing now? Early ones were capable of performing only repetitive and primitive tasks. But nowadays they are getting more and more sophisticated. Machine Learning has changed the nature of bots completely: now they are able to learn from the content they are operating with, becoming more and more complicated. Contemporary “androids” are able to mimic human behavior in social networks in such an accurate way that it’s not so easy to realize whether it is a bot or not (numerous Instagram post bots demonstrate it well). Moreover, they complete data research or even trade on a financial market. Basically, researchers tend to separate two big groups of them: so-called “Good” and “Bad” bots. The first one is completely non-threatening and useful for online users. These are:
- Monitoring bots. They are used to check the availability and correct functioning of different websites and other online features
- Commercial crawlers. These bots are collecting data on behalf of digital marketing tools
- Search engine bots. This type of bots is made to extract information for search engine algorithms
- Feed fetchers. Their aim is to collect content to mobile and web applications
But, how it usually happens with technologies in the history of humanity, there are some malicious ways bots are being used. So, speaking of Bad Bots, there are different types of them:
- Impersonators. These ones are taking false identities to come down security features. They are the most common among the bad ones
- Scrapers. That’s the name for the bots which aim is to gain unauthorized access to content and publish it somewhere over the internet
- Spammers. the title says it all: these bots are created to propagate junk content everywhere they can. For example, have you ever noticed the strange Instagram users, who are leaving primitive and preposterous comments (“Damn!’, “Cool!”, etc.) under your posts? These are low-quality Instagram comment bots -- iconical representatives of the spammer.
- Hackers. The mentally disabled evil brothers of a monitoring bots. These guys are doing almost the same -- looking for some vulnerabilities of websites -- but in order to use it for data theft malware injection, etc.
How can I use them?
Messenger bots are to simplify your work with different services by giving a mean to do it directly through messengers (they are widely-used in Facebook Messenger and Telegram). It is a great tool to minimize the time you are usually spending on shifting from window to window or from one app to another. With the helping hand of these developments, you can get the information you need just by using keywords. The most common types of these bots are informational -- you can ask them about the weather, currency rates, etc. Also, they are widely spread among the internet-shops.
If you are not affiliated with hacking, don’t hold any website or not interested in solving the great concern of artificial intelligence, you can still find bots useful. Whatever you are doing, you have to promote it. The most effective and save-proven way to do it is via social media. People are recently getting obsessed with socials: an average user spends about 2 hours a day on them.
Instagram has become one of the most fruitful platforms for promotion. The reason for its fast-growing popularity among the business is not so evident: whether it is its centration on visual content, the interface simplicity or the variety of functions it is providing. Despite all the advantages, Instagram has one big drawback: its competitiveness. The dozens of content and “info-noise” there makes Instagram promotion significantly effortful activity. The desire to automize it is completely natural: neverending grind with likes, follows and comments isn’t worth your time. But it is the kind of tasks bots are perfect in: repetitive acts in a well-defined narrowed sphere demanding almost no cognitive agency. Why not to benefit from it?
So, what are Instagram bots capable of:
- Instagram like bots are the most simple ones. They just automatically like different posts all over the Instagram, so that somebody of the affected by their activity may get interested in the “person” who enjoys his/her posts. This type of Instagram automation is the most primitive and not so effective as the next options.
- Scheduled posting. This feature is essential for intensive Instagram growth. You have to keep posting on the regular basis to keep your audience engaged. Unfortunately, it is not always possible: different time zones, any kind of aleatoric events which interferes your capacity to post or just your desire to get rest of socials may break the content plan.
- DMing your target audience. It is a great feature of any social media -- the ability to address directly to anybody, whether it is an individual or a major brand. But it is impossible to chat with thousands of people you are interested in. Here’s where bots come to your help. Of course, these ones will never pass the Turing Test, but at least they are capable of informing your audience of the most significant news you have.
- Tracking comments. Bots provide you with the opportunity of collecting all the comments in one place and analyzing them with different tools. It significantly streamlines your work with comments.
Is it safe to use this bots?
Messenger bots are 100% secure. Instagram bots are a little bit more tricky. If the bot you are using is third-rate, you will get banned for using it. Instagram has its own algorithms to detect fake activity. Mostly, they are based on some limits of activity frequency. Well-made bots usually mind this factor and don’t act too fast. What is more, spam bots are getting banned instantly, so it is better not to use Instagram comment bots. But anyway, it is always up to you to use it or not.
What to expect from bots in the future?
Since there is too much noise around Machine Learning, it is getting hard to understand what to believe. The enthusiasts are extremely optimistic: you can find tons of articles about how AI will replace humans in any sphere and change the world completely. In the end, it is almost impossible to predict the way technology is going be develop, at least we can say that the solutions bots are giving now are worth using.