AI in 2024: State Of The AI World (Part 2)
Part 2 in on our snapshot look at AI in 2024 focuses on adoption trends this year and in my personal use
Posted by Charlie Recksieck
on 2024-07-18
Click here to read Part 1
* * *
Actual progress in 2024
Already in 2024, we’ve seen rapid enterprise adoption of generative AI (GenAI), significant technical improvements in reasoning and smaller models, a surge in private investment, and an intensifying focus on regulation and governance. AI transitioned from an experimental technology to an essential business tool embedded in everyday life.
Risng Adoption: Approximately 78% of organizations reported using AI in 2024, up from 55% the previous year. GenAI adoption in particular nearly doubled in one year, with 65% of companies using it in at least one business function, such as marketing, sales, and IT.
Performance Leaps: AI models have definitely demonstrated improvements on
complex reasoning and programming tasks, in some cases outperforming humans in limited time scenarios. This was driven partly by advancements in training techniques and model architectures.
Emergence of Agentic AI: A major trend was the rise of AI agents designed to operate autonomously, plan, and execute multi-step tasks to achieve specific goals without direct human oversight. Companies like Salesforce and Google released platforms for building these agents.
Multimodal AI: AI systems capable of processing and integrating multiple data types (text, images, audio, video) became more sophisticated, enabling a wide range of new applications in areas like healthcare diagnostics and real-time translation.
AI in Science and Healthcare: AI’s impact in scientific research is already profound, leading to faster drug discovery, improved genetic diagnostics, and breakthroughs in areas like nuclear fusion research and wildfire detection. The FDA approved a record number of AI-enabled medical devices.
Record Investment: Private investment surged, with the U.S. leading globally, followed by China and Europe. Businesses are increasingly moving AI from innovation budgets to permanent operational budgets.
And then there's how things have transformed the workforce. AI can automate repetitive tasks and boost productivity. Companies say they are investing heavily in upskilling and re-skilling programs to prepare employees to work alongside AI technologies.
But let's say that AI can reasonably shave off about 7 out of 10 staff workers at a software company, just as one example. I've discussed this with friends and overall that's about the number we're landing on for percentages of current programmers who soon can be replaced with very solid automated code generators. That future is almost here. Go ask ChatGPT to write something for you; it might not come out absolutely 100% done and ready for market, something I need to write a blog post about soon.
Anyway, let's say that automation can get rid of about 7 out of 10 coders. Yes, you will have to bring on 1 A.I. specialist to manage and integrate A.I. content. That's 1 new job out of the 7 of 10 who can be replaced. The argument for AI specialists being a huge lift for the workforce is pretty flimsy.
In manufacturing it's even worse. Amazon warehouses used to employ and amazing amount of people. But as those duties become automated, you might need ONE AI/machine specialist for the automated replacement of every FIFTY workers.
My Personal Uses
Writing - When it comes to writing something like, let's say, a blog - hopefully most of what I have to say is from original thoughts, an outline and trying to make a certain point. But when it comes to make intelligent arguments about complex developments in technology - sometimes it's great to have ChatGPT handle the boring research part and hope that it's getting its facts right. For example, completing a text prompt of "write 200 words on how artificial intelligence has affected medicine" or something like that saves me the research time. Then I clean up the language into something that sounds more like me. That's the labor-saving idea. But are the facts that I'm getting from ChatGPT all 100% correct? Not sure. Is this definitely not plagiarism? Not sure.
Art Work - My fun weekend band, The Water Spots, has been making trippy posters from AI image from text prompts. Take a look at our poster gallery - Additionally, the image at the top of this article was generated by a "generate-image-from-text" AI prompt.
Code - Several sites, even free ones, are doing a great job of writing code from a text prompt like "generate web calendar with holiday in Python". I'll get into this in greater detail on a blog post here soon about how AI not only does the gruntwork but also a terrific way to teach yourself new programming language.
Going Forward
And AI is in all of our lives at present in 2024. Improved Google searches, smarter directions on phone, better Spotify and Netflix suggestions, and that's just the trivial stuff. Maybe AI technology was critical in getting you a recent medical diagnosis that your old family doctor would have missed.
Hang on to your hats. I'll guess we'll all have to see where this is going.
