Bill Skelly
AI, Data Analysts, Solutions Review
August 14, 2023
This was first published in Solutions Review.
Regardless of your profession, we’re all familiar with the triple constraint of speed, quality and price in business.
Conventional wisdom says we can have only two at the most. If the priority is high speed and high quality, you’ll pay a high price. If you want good quality at a good price, then speed will suffer.
Forward-thinking data analysts and consultants specializing in AI, machine learning, predictive and prescriptive modeling will advise their clients to prioritize quality and accuracy over speed and price.
Unfortunately, AI technology innovators are scrambling to be the first at the cost of everything else. In the new space race for AI dominance, technologies are being released before they’re ready for prime time.
Case in point, Auto-GPT, is an experimental open-source interface to GPT-4 and GPT-3.5, which can complete tasks autonomously. Compared with OpenAI’s ChatGPT, Auto-GPT automates multi-step projects that would have required many prompts for ChatGPT to accomplish. In addition, Auto-GPT can access websites, search engines, apps, software and services—both online and local. The ability to gather external data allows Auto-GPT to self-evaluate, verify collected data, remove inaccurate and sub-par information and create a new task to get better data.[1]
Max Tegmark, cofounder of the Future of Life Institute and a leading AI Safety researcher at MIT, says these types of experiments are violating safety norms for AI development, such as:[2]
The Future of Life Institute recently posted an open letter to “Pause Giant AI Experiments” which received more than 11,000 signatures, including Elon Musk, Steve Wozniak and Tristan Harris of the Center for Humane Technology.[3]
We need to keep in mind that even with the most bleeding-edge AI iterations, they are simply tools. The technologies lack critical thinking and ethical judgement. Humans must stay in the driver’s seat to harness the power of generative AI and machine learning.
There’s no denying the value that data scientists can derive from AI, machine learning, large language model (LLM) and natural language processing (NLP). Machine learning is behind the big data analytics each time we get personalized recommendations from Netflix. Google’s predictive text in emails and optimized directions in Google Maps are also being powered by machine learning.
Starbucks’ Digital Flywheel program merges digital and physical customer interactions around rewards, personalization, payments and orders. Using big data, AI and NLP, Starbucks delivers highly personalized emails based on customers’ past purchases. Instead of the typical few dozen emails sent monthly with offers to the broad Starbucks audience, the digital flywheel produces more than 400,000 personalized weekly emails featuring different promotions and offers.[4]
Other ways LLMs and NLP can be used by data analysts include creating code and applications to analyze information and automate processes for gathering, formatting and cleansing data. The tools can define the charts, diagrams and infographics to be included in reports, as well as offer guidance on compliance and regulation to make certain data operations are legal, unbiased and ethical.[5]
Given generative AI’s fast-paced advancements and ongoing massive layoffs at leading firms, it’s understandable that some data analysts may be concerned about security in their career.
If AI can collect and analyze data within minutes, plus recognize patterns and compile the information into formats that are easily understood by colleagues and clients, are data jobs at risk?
The bottom line is that even sophisticated LLMs and NLP tech can’t solve complex problems. They don’t have critical thinking and strategic planning capabilities. So even as AI becomes more established and gains widespread accessibility, data scientists, analysts and consultants will continue to be vital to long-term business success.
Here are three tips for staying up to date on the latest AI innovations:
1. Assign a team to integrate machine learning into your organization’s standard operating procedures that will benefit clients.
2. Make certain team members embrace lifelong learning, receive regular training and keep certifications updated.
3. Invite data analysts to attend webinars and industry events to ensure your organization keeps up with the newest trends.
By staying open-minded and open to learning, data analysts can harness AI technology to be more productive while maintaining quality and affordability.
William Skelly is CEO of Causeway Solutions, a leading provider of Acquisition Analytics and innovative data services. Bill serves as advisor with some of the nation’s most influential organizations—from grassroots public affairs efforts to U.S. Presidential campaign strategies. Causeway Solutions empowers clients to make smart, timely, data-driven decisions through real-time consumer insights to better reach target audiences.
[1]: “What is Auto-GPT And Is Now The Time To Freak Out About AI?” Roger Montti, Search Engine Journal
[2]: “Summary: The Case for Halting AI Development – Max Tegmark,” Effective Altruism Forum
[3]: "Pause Giant AI Experiments: An Open Letter," Future of Life Institute
[4]: "How do Big Data and AI work together," Kathleen Walch, TechTarget
[5]: "Will ChatGPT Put Data Analysts Out of Work?" Bernard Marr, Forbes