If on the one hand the search forArtificial Intelligence meets the limits, due to the increasing cost of research, on the other hand more and more companies are trying to expand its use thanks to the possibility of using technology without the need for planning. It is the “no code” phenomenon.
Just as digital photography has given everyone the opportunity to make sophisticated modifications to their photos, so does codeless artificial intelligence make dozens of applications available to create advanced tools that use the capabilities of artificial intelligence.
Fast learning for AI: what are the three most promising techniques
The parable of artificial intelligence in technological innovation
In the field of artificial intelligence we are witnessing a rather classic parable in the field of technological innovation: the innovative idea takes its first steps in research laboratories, where specialized people study its possibilities and fields of use. Innovation is then designed and its dissemination begins to a wider audience of techno-enthusiasts and among those companies that have the resources or innovation capacity that allows them to start using it. The last step finally sees technology evolve in that direction increase usability more than its potential, thus making it usable by individuals – individuals and companies – who have only a limited knowledge of the technology itself, thus ensuring a wide spread of innovation. In short, it’s the same move illustrated in my previous article, “Quick Learning about Artificial Intelligence: What Are the Three Most Promising Techniques,” which describes the technologies that try to make artificial intelligence available even to those organizations that do not have the resources (skills, data and time) required to develop effective artificial intelligence from the beginning.
A deep learning guide: what it is, how it works and what the application cases are
This path to the widespread use of artificial intelligence is paying off: a growing number of companies are offering simple and easy-to-use artificial intelligence services that allow even non-experts to apply this technology without having to write a program bar: very often all you need is a browser and enough ingenuity.
The various codeless AI applications
Microsoft-based company Lobe.ai offers, for example, a free application that lets you use your computer camera to automatically capture, tag, and create systems for recognizing certain types of images with minimal interference. , above all, without requiring in-depth knowledge of the underlying algorithms. The demo video shows the creation of an artificial intelligence that can recognize whether the person in front of the computer is drinking or not, made in a matter of minutes.
For example, the article “No-Code Brings the Power of AI to the Masses,” published by the New York Times last March, describes the work done by a beekeeper, which has created a system capable of recognizing insects entering its hives, thanks to a microcamera installed at the entrance to the hive and a “codeless” artificial intelligence system from Lobe.ai, which “taught” to recognize a bee. Thanks to the use of artificial intelligence, our beekeeper was able to quickly identify, among the thousands of photos taken, those that did not contain bees, thus limiting the number of photos that will be analyzed and locating the “intruding” insects.
Lobe.ai not only runs it artificial intelligence training, but helps the user to create images that will be used for learning. Once the learning is complete, it is possible to check the effectiveness of the recognition and possibly “correct” the effectiveness of the recognition simply by pointing out the cases in which the image has not been sorted correctly. Once the learning is completed with satisfactory results, the network can be downloaded in different formats and then used in other applications.
On its page dedicated to possible uses, Lobe.ai states, in addition to the identification of the bees mentioned, and the identification of the plants, the counting of repetitions of gymnastic exercisesThe fire surveillance, to name a few.
But Lobe.ai is not the only company operating in the field of artificial intelligence “without code”: Jujifor example, it is a tool designed to make chatbot creation as easy as creating PowerPoint presentations. Juji uses machine learning to manage chat flows and recognize specific user characteristics so that it can personalize human-machine interactions instead of just producing interactions that are the same for everyone. Thanks to predefined chat routes and selectable chatbot “personalities”, even a person with very superficial artificial intelligence knowledge can quickly create an interactive chatbot can help the user perform a wide range of activities, such as welcoming users to a site, responding to user requests for online offers, and guiding the user’s actions in searching for specific information.
Codeless development platforms
Codeless AI products are part of a much broader movement, the No-Code Platform Development (NCDP), which aims to enable developers and non-developers to create application software through graphical user interfaces. . The popularity of these development platforms is growing, thanks to the ability to accelerate application development and due to the limited availability of dedicated developers.
Obviously these are products that have their limitations: Like all artificial intelligence systems, their effectiveness depends on the quality and quantity of data available and many factors can reduce the effectiveness of the result. For example, in the case of recognizing images in different lighting conditions or with different backgrounds, once you leave the simplest scenarios or projects, the success rate may remain below the expected results, unless you have a large number of examples related to each of the cases.
However, the expectations of those working in the industry are very high, as codeless artificial intelligence tools, which are part of the wider NCDP movement, are driven by the same needs. In addition, the possibilities offered by machine learning are still poorly understood and explored, and therefore the prospects and possibilities of “codeless” artificial intelligence are expected to increase as more people understand its possibilities and uses.
GPT-3 and DeepMind
In addition, recent developments in the capabilities of artificial intelligence systems are intended to further facilitate the development and usability of codeless platforms: for example the GPT-3 text creation system, created by OpenAI, a company co-founded by Elon Musk, is able to write program modules following the expressions expressed in natural language. Thanks to these features OpenAI was able to create GitHub Copilota tool that works as an autocomplete feature for developers, making their work faster.
DeepMind, a subsidiary of Alphabet, Google, has gone a step further with an artificial intelligence tool that can write complete code to solve complex problems posed by voice or text commands. Its users Power Platform from Microsoftwhich includes a family of codeless products, can create applications simply by describing the function they want to achieve.
These developments suggest that the time is approaching when it will be possible to generate a report or export data using only natural language, for example: “Search all client files from the last year”. This possibility could lead to a drastic reduction in office work, leaving employees more time to focus on activities with higher added value.
Right now, however, the majority of codeless system users are still the professional users of the companies they want. simplifies tasks without necessarily involving a programmer. The next step, in fact already in progress, will be to get these tools into the hands of non-professional users and this will open the doors to a very rapid increase in the number and variety of uses.
@ALL RIGHTS RESERVED