Trac Bannon's AI Predictions for 2024

Posted by Tracy Bannon on Sunday, January 7, 2024

๐— ๐˜† ๐Ÿฎ๐Ÿฌ๐Ÿฎ๐Ÿฐ ๐—”๐—œ & ๐—ฆ๐—ผ๐—ณ๐˜๐˜„๐—ฎ๐—ฟ๐—ฒ ๐—”๐—ฟ๐—ฐ๐—ต๐—ถ๐˜๐—ฒ๐—ฐ๐˜๐˜‚๐—ฟ๐—ฒ ๐—ฝ๐—ฟ๐—ฒ๐—ฑ๐—ถ๐—ฐ๐˜๐—ถ๐—ผ๐—ป๐˜€?!?!

Hello friends, itโ€™s been a long timeโ€ฆโ€™23 was quite a rollercoaster for me personally and professionally. I am so thankful for the support of my network, my family, and my friends. I am ready for โ€˜24!

The question is, are you tired of hearing about AI yet? Iโ€™m personally tired of the deluge of generated blog posts with all the embedded unicode emojis. That said, 2024 will be more exciting, innovative, and possibly dangerous than 2023.

๐Ÿญ - ๐—”๐—œ & ๐˜๐—ต๐—ฒ ๐—ฆ๐——๐—Ÿ๐—–

The biggest impact on me day in and day out is applying AI to Modern Software Practices and completely changing the software development lifecycle. This is so much more than code generation!! Imagine capturing what end users want with natural language and requirements being generated complete with suggestions for where there are gaps?! What about having AI provide you with an analysis of alternatives or tradeoff analysis? Shazam!! This is more than game changing, gang. The SDLC as we know it may become unrecognizable in the coming months and years. Iโ€™m laser focused on tooling myself to be an ๐—”๐—ฟ๐—ฐ๐—ต-๐—”๐—œ-๐˜๐—ฒ๐—ฐ๐˜!

๐Ÿฎ - ๐— ๐—ผ๐˜ƒ๐—ถ๐—ป๐—ด ๐—ณ๐—ฟ๐—ผ๐—บ ๐—ก๐—ผ๐˜ƒ๐—ฒ๐—น๐˜๐˜† ๐˜๐—ผ ๐—ฃ๐—ฟ๐—ผ๐—ฑ๐˜‚๐—ฐ๐˜๐—ถ๐˜ƒ๐—ถ๐˜๐˜†

There is still a ton of novelty around Generative AI in particular. Moving into the new year, we will see much more rigor by organizations evaluating AI and generative technologies. As the novelty fades, the imperative shifts to demonstrating tangible productivity gains and new value.

This is not just about adopting technology for its own sake but about integrating it in a way that enhances operational efficiency and drives innovation. The crux here is not just adoption but adaptationโ€”understanding how these technologies can be tailored to specific organizational needs and goals.

Adopting AI comes with its set of challenges. A significant one is the initial loss of productivity. For example, what if it takes a marking specialist 2 hours to engineer the perfect prompt when they could have written the copy in 30 minutes? If they are able to reuse that prompt, it will have a return on investment. What about the situation when an early career software developer generates code for a function? Generally they are presented with 3 or more options to choose from. They often donโ€™t have the skill to discern which option to pick and worse, code generation is still in its infancy with many defects and gaps in the code.

๐Ÿฏ - ๐—•๐—ฒ๐—ถ๐—ป๐—ด ๐—”๐—œ-๐—ฆ๐—ฎ๐˜ƒ๐˜ƒ๐˜†

In 2024, we will see continued security breaches by the average employee. The cause? Lack of adequate training. That leads me to my next prediction, 2024 will be the year of being “AI savvy”. Being able to use AI tools relevant to your career will be as essential as being proficient with traditional office productivity tools.

2024 isn’t just about knowing how to use the technology; it’s about understanding its implications, strengths, and weaknesses. Without widespread literacy in AI and ML, organizations risk not fully leveraging the technology’s potential and exposing themselves to avoidable vulnerabilities.

My Predictions?

๐Ÿฐ - ๐——๐—ฒ๐—บ๐—ผ๐—ฐ๐—ฟ๐—ฎ๐˜๐—ถ๐˜‡๐—ฎ๐˜๐—ถ๐—ผ๐—ป ๐—ผ๐—ณ ๐—œ๐—ป๐—ป๐—ผ๐˜ƒ๐—ฎ๐˜๐—ถ๐—ผ๐—ป

That leads me to my next prediction for 2024: AI will become increasingly accessible and we’ll witness an acceleration in innovation and the birth of new ideas. Iโ€™m going to call this the democratization of innovation. More people than ever can contribute to the technological landscape. Advances in Low Code and No Code platforms allow there to be citizen developers. . It may sound tacky to say: with great power comes great responsibility. This democratization also introduces significant risks, especially around security. As more individuals gain the ability to manipulate and implement AI, the potential for misuse and mistakes grows.

The wave of innovation canโ€™t be stopped but we can focus on security and ethical considerations.

๐Ÿฑ - ๐—ก๐—ผ๐˜ƒ๐—ถ๐—ฐ๐—ฒ ๐——๐—ฒ๐˜ƒ๐—ฒ๐—น๐—ผ๐—ฝ๐—ฒ๐—ฟ๐˜€

My next prediction for 2024 is the rise of novice developers. We used to use the term “script kiddies” to mean self taught hackers. Being self directed is an inevitable trend. I may sound like a broken recordโ€ฆ 2024 must be the year of AI education and guardrails. The challenge, of course, will be striking a balance between fostering innovation and maintaining quality and security rigor.

๐Ÿฒ - ๐—”๐˜‚๐˜๐—ผ๐—ป๐—ผ๐—บ๐—ผ๐˜‚๐˜€ ๐—”๐—ด๐—ฒ๐—ป๐˜๐˜€ ๐—ฎ๐˜€ ๐—ง๐—ฒ๐—ฎ๐—บ ๐— ๐—ฒ๐—บ๐—ฏ๐—ฒ๐—ฟ๐˜€

From a pure tech perspective, I anticipate the roll-out of autonomous AI agents. Right now, AI is often a tool for us to use. As it is further integrated into our daily lives, it will act without the same level of prompting and perhaps, without it. These agents, capable of operating independently and intelligently, will redefine interactions and operations across various sectors

๐Ÿณ & ๐Ÿด- ๐—•๐—ฒ๐˜†๐—ผ๐—ป๐—ฑ ๐—Ÿ๐—ฎ๐—ฟ๐—ด๐—ฒ ๐—Ÿ๐—ฎ๐—ป๐—ด๐˜‚๐—ฎ๐—ด๐—ฒ ๐— ๐—ผ๐—ฑ๐—ฒ๐—น๐˜€

2024 will see emphasis on two different model types: multi-modal model and small language models. Multi-modals understand and generate information across different types of data. The outcome may be a more integrated and nuanced understanding of the world.

Smaller, focused language models In 2024, smaller, more specialized models will be tailored to specific domains. 2023 was all about large language models, LLMs, a sort of one-size-fits-all. Itโ€™s time to focus on precision and efficiency. Weโ€™re seeing bespoke models customized to the nuances of different sectors like defense, fintech, transportation, and healthcare. A value add with these models beyond enhanced performance is mitigating some risks with broad and generalized models.

I am over the moon excited at whatโ€™s to come. We all need a mindset that balances innovation with responsibility, agility with security, and ambition with ethical consideration.

The future is not just about what technology can do but how we choose to use it.

My hope is to dive into each topic a bit deeper with yaโ€™ll in the coming days. ๐—ฌ๐—ผ๐˜‚ ๐—ด๐—ฎ๐—บ๐—ฒ?

(Pix thanks to DALL-E.)

#SomethingToNoodleOn #SoftwareArchitecture #GenerativeAI #DevSecOps #GenAI #AIML