Our vision

Generative AI is constantly progressing, both in cognitive abilities and cost-effectiveness, leading to significant productivity gains. Inevitably, the software industry is moving toward an era of abundance, where professionals can produce more than before, and individuals are beginning to create as well (quality versus quantity).

​

This abundance could shift certain paradigms (will customizable/integrable software be able to compete with AI-boosted development from scratch?) and, most importantly, increase demand. At the same time, these gains will allow IT professionals to focus on differentiating elements: code quality, the robustness of software architecture, and the quality of conceptualization and specifications.

​

This abundance is already raising expectations for these differentiators (similar to consulting firms, which now must produce deliverables for support that used to be at the level of final project deliverables). Furthermore, this abundance highlights the need for intellectual property protection.

​

Meanwhile, the development of multi-agent systems is leading us to rethink the relationship between humans and software. By keeping humans "in the loop," they can secure and supervise AI assistants' actions on certain applications.
Moreover, as humans and generative AIs now have comparable cognitive abilities (for certain tasks), it is necessary to reimagine IT systems to efficiently integrate these agents and automate as many tasks as possible, leveraging the concept of distributed cognition.

Some experts are even beginning to speculate on "living software," capable of self-improvement through user feedback and navigation data. This principle is no longer a distant dream but a long-term achievable goal. At Swoft, we have designed our framework around the principles of "Specify, Make, Adapt" to eventually empower the applications created from it.

​

In the face of these challenges, we believe that the success criteria for systems will be:

• Quality of design

• Providing context in software to integrate AI agents

• Synchronizing data capture from all sources

• Documentation

• The ability to create empowerment rather than sidelining professions and their practices

• Intellectual property protection and the possibility of decentralized exchanges

• Implementation/Mastery of software with robust and modular architecture (for critical and complex systems)

• The ability to elevate the most junior profiles, making them more competitive than AI models

• The ability to track expected productivity gains with Gen AI.