I’ve recently realized that AI is inevitable. It will change the world. Quickly. Within the next 2 years. Any organization not exploring how to use AI to enhance their business now will not exist in 3-5 years. They will not be able to compete.
The traditional role of the “analyst” job category will morph into, WHO knows how to formulate queries to the AI that will return the desired results. There will also be an analyst/technologist’s role in evaluating WHICH AI on the market is best to use, and also, how to develop an in-house AI — sourcing the hardware and the material used to train the AI. Turning the AI lose on the internet to find it’s own training material will be very risky.
I have followed Mike Adams’ venture into the world of AI. He has developed his own Large Language Model that can answer queries about health, nutrition, sustainable farming - those types of subjects. Many leaders in these fields are giving him their libraries, both digital and paper, to feed into the AI during its training period. (People like Sayer Ji and Joe Mercola are donating their material to Adams.) Enoch is a Large Language Model (LLM) currently in development, that will be available to download this spring. This LLM will run locally, on your desktop, without any connection to anything in the cloud. I have no idea what the technical specs to handle this application will be.
Right now, there are job opportunities to develop in-house AI LMMs. If you feed the LMM content, it will regurgitate an analysis and summary. If you give it a book, it will read it to you. OR give you a summary if you prefer. It will create graphics and movies if you prefer the visual arts. Users will merely query what the developers created.
The podcast linked below talks about an announcement from nVidia - this spring, a $3000 piece of hardware will be released that will replace $150,000 servers currently used to develop and run AI LMMs. This will change everything in this industry, and empower virtually everyone the ability to dabble in the world of AI LMM development. This technology will be open sourced and will achieve decentralization of AI LMMs. Content known to the LMM will no longer be in the hands of Microsoft or Google.
This is a ground floor opportunity for teens and 20 somethings embarking on their journey in the world of work.
I sat on the edge of a technical revolution in the late 70s/early 80s, when computing technology was poised to take a leap into expanding human capabilities. In 1976, I was introduced to full text searching using Lexis in the law office. These were “dumb terminals”, hooked over phone lines to a mainframe computer. There was no internet, and no IP technology. Most all was still analog. I learned how to formulate queries to return what I was looking for, and only that. Back then, this method of legal research was SO expensive, you had to have knowledge of which legal library to use to perform the query at hand - failure to do so got you called into the office to explain the thousand dollar bill resulting from your search. Today, I am still amazed at how many people do not know how to formulate a query, or how to determine where to look to return relevant and high quality/vetted search results. I also was instrumental in moving computers onto the desktops of everyone in the organization, and teaching them how to use them. I also supported hardware and software issues that inevitably developed.
See the vision and prosper!! My insights into where the technology was going served me well in securing well paying and interesting employment. This is a huge opportunity for many who can connect the dots, see how the work world is changing and turn this knowledge into job skills.
https://www.brighteon.com/ee799280-bba2-43fa-ba2f-9ed0d8d775b0
Post script:
https://open.substack.com/pub/corbettreport/p/2025-predictions?r=1736yu&utm_medium=ios
James Corbett links 1/1/2025 AI insights, including one on how LMMs can and do lie to achieve their objectives (- but where do their objectives come from?) He also links many other insights into how AI is developing and who has plans for using it how.
No comments:
Post a Comment