What's surprising when considering generative AI (gen AI) is its impact on the IT community. A year ago, the belief heading into this was that next-generation AI would spark the next boom in data science. It seems like the opposite. It looks like many of the data science and data management functions will be democratized.

For example, companies were putting a lot of effort into labeling their data. But gen AI can label data very easily. Perhaps the most important aspect of this democratization is that once LLM is deployed on data, the average employee or executive can examine the data and perform analyzes that were previously the norm for data scientists. It means that it becomes.

While LLMs can integrate disparate data from many sources and uncover patterns and insights using often incomplete or inconsistent data, they place this power in the hands of the average knowledge worker. The fact that you can do this essentially democratizes the role of data science.

It's not that we don't need data scientists anymore, it's that we almost eliminate the need for data scientists.

As data science becomes more democratized, data science capabilities are pushed out into businesses. Even at this relatively early stage of the AI ​​generation, we are already seeing a decline in the need for large data science teams.

Additionally, while the most sought-after skills learned in college once guaranteed multiple jobs, we are now seeing new graduates struggle to find positions. This suggests that the need for these hard-to-learn skills could change significantly, and the average pay for this role could drop significantly. For third-party consultancies that specialize in this, this also signals a slowdown in demand and the need to diversify quickly.

On the bright side, gen AI will create an entirely new job title: Prompt Engineering. Although this new role is still in its infancy, it is likely to utilize similar skills to data science and could provide a career path for those with data science training and inclinations.

Like all technologies that have come before it, Gen AI will disrupt existing roles and careers while simultaneously creating new ones. The Prompt Engineering role seems to be one such example.

Gen AI seems to have a similar impact on coding development, but to a lesser extent. There are many professional gen AI tools, such as GitHub, that can write your code or greatly assist you in writing your code. As these tools mature and the programming and engineering communities adopt them, we can expect effects similar to those currently observed in data science.

From writing code, testing code, and debugging code, gen AI tools appear to greatly increase the productivity of programming and engineering teams, allowing them to run code much faster. However, as with data science, while some sets of activities have been automated and others have become dramatically easier, the need for more advanced skills in architecture and design will remain, and perhaps may increase.

At the moment, there is a big debate about whether companies will need fewer engineers as genetic AI matures and is deployed. The argument is that significant increases in productivity will reduce the number of engineers needed and create a surplus of skilled labor. This will result in lower wages, less job creation, and less need for a significant number of engineers to find alternative careers. If this scenario were to play out, it would have the greatest impact on third-party service providers and could cause significant disruption in countries such as India, where large numbers of these engineers are present to meet existing demand.

However, another scenario argues that as the cost of engineering declines, the demand for programming and engineering will more than offset productivity gains. Under this scenario, we would expect the need for engineering to increase with rising wages as workers capture a portion of their own productivity.

I personally support this alternative scenario and point to the history of information technology as a likely guide. Whenever an innovation occurs in information technology that lowers the price of the technology or the cost of implementing and maintaining it, the demand for more technology quickly outweighs the reduced unit price, resulting in an expanded market and higher wages. Did.

To see this phenomenon, we don't need to look back just at cloud migration. The cloud has significantly reduced the unit cost of data processing while simplifying many management functions. However, this did not lead to a reduction in IT spending as demand increased rapidly and costs fell. As a result, overall IT spending has increased.

In a possible scenario where technology becomes more accessible and cheaper to create, deploy, and maintain, companies could double down on their technology investments and the need for programming and engineering talent would grow even further. This talent is also likely to require more analytical and integrative skills, which will likely require skill upgrades and lead to higher wages.

The impact on consulting, systems integration and outsourcing companies may be more subtle. It is clear that this will cause disruption to existing models and pricing structures. The prices that can be charged for the construction and maintenance of existing properties may necessarily decline. Managed services will need to be re-established and relationships will likely need to be re-established.

This may result in enterprises consolidating workloads to fewer service providers. Additionally, companies that outsource existing large-scale facilities may want to implement some or all of these functions as they build or expand global in-house centers (also known as captives) in low-cost destinations such as India. There is also the possibility of encouraging in-house production.

However, increased demand for information technology will more than offset these headwinds. As with individuals, they disrupt the existing status quo and challenge existing business models, but they also have the potential to open up new, more profitable services and business models.



Source link