Webinar State of AI: 2024 with CodeTraining
The presentation was done by Markus Egger, President and Chief Software Architect (among many other titles) on the current status of AI and how it is available for business development in archival response in not only customer settings but with employee registrar queries and workload tasks.
AI is a feature working interchangeably with other models in text and image. As the big competition markets with China and US where India's space program (no doubt near this as well) takeoff are focusing on their future investments and where we lack in robotic performance (in comparison to China), we are aiming for AI goals instead.
Economical Impact with AI and Competitive Business Goals with AI
"The expected characteristic is that mainly the "middle" will be affected, while low-skill and high-skill jobs should be around longer" made an impact on visionary statements for employment, resource, sustainability, and competitive business strategies for investment. It is still a risk, more research, development, and preference for determining how we grow as entrepeneurs to the enterprise levels into government policy-making roles and how it will affect our leaders decision-making, education expectations/training, into the workforces they provide.
All Data Sets (Freely Available), Availability of Compute Power (Cloud services, platform service arrangements), Funding and Entrepeneurship, Well-trained AI Scientists, A Supportive Policy Environment.
The AI term was coined in the 1950's, envisioned in usage that is not referenced today in the same light. It has shifted into what is now possible, available, and improving with each passing architectural framework into a "factful" Retrieval Augmented Generator (RAG) provided by programmer discipline. AI is not emotional but it will be able to talk to you in a relational interpretation, however, you must be capable of fact-checking and realizing that this is still new. GPT3 was released in 2020 and a year ago (CodeTraining's first webinar on the State of AI) Microsoft's Co-Pilot AI was not available. GPT4 recently released in its latest version in 2023 while other teams are also venturing in their competitive architectural models like Stable Diffusion 3 (An image generating model that has caused errors in censoring and filter errors in competition with TensorFlow with Machine Learning models), Whisper, Gemini, Mistral, and more.
How can NEW-TRG use this informational review?
Global marketing, global competition, self-sustainability in business, and educational funds in AI, IT, Machine Learning backgrounds are highly encouraged. The future of the world is separating at a fast pace and with AI, is only going to boost performance and speed of work task flows, with less cost and less long-term damage or error (human-made or not). The accessibility and strive for improvement is now.
What AI encounters and the knowledgeable user may encounter...
I recently had an encounter doing ASP.NET MVC (Model-View-Controller) homework and asked Co-Pilot a question on PartialViews, that had an error in the descriptive language it chose to portray in the usage of the initial question. I corrected it and it acknowledged its mistake by supporting my correction. I was surprised by the response, and how in the Webinar, stated how "factfulness" developers strive to make AI tools. I became skeptical and other questions/concerns arised:
1. What imagery it cannot comprehend or relate in its responses - grabbing from multiple sources - yet can create imagery in some models... it is going to bridge this connection at some point, the assumption is neural networking models.
2. How the AI model could be corrected in a basic term and function with a well-known software program that I was searching for syntax in my question.
3. Where knowledge in these tools, data, and concepts are needed and should be documented for improvement in not only the AI but how we are recruited as developers during this time. I thought of it as a 'while I can' since AI is still in a young age with the world.
4. Other models with ChatGPT would ask the same question and the output would change, deriving multiple sources, reflecting average or library source or data governance flaws to its response factfulness being degraded. This was not the case, since it repeats these basic concepts and examples to not reveal answer but now act as an educational tool.
Conclusion
Technology is young, we can definately step in and learn now.
Competition in software engineering, architecture, and the cost impact depends heavily on the programming team and that's not mentioning the hardware and memory space with server and networking dilemmas to add to the chaos of what is a work space in today's working world!
The data being collected on you, your business, government, and global affairs. It's time to join the wave of technological input!
