From Artificial Intelligence to Human Intelligence

Introduction

Human intelligence has entered a transformative phase with the latest advancements in AI, particularly generative AI. This technology serves as a catalyst, bridging gaps in processes and seamlessly connecting human interactions with underlying technical systems. The success of generative AI, however, hinges on context and data quality for effective decision-making and outputs.

Forward-thinking HR and recruiting teams have meticulously curated a wealth of structured data related to performance, engagement, and well-being. This contextual data serves as a foundation for both human and AI actors. In this era, where engagement and productivity are (re)defined as interdependent, generative AI finds applications in boosting productivity and enhancing engagement.

This guide explores the transformative journey of HR practices, focusing on the collaborative efforts between the Society for Human Resource Management (SHRM) and Humanly. In this guide, we will also highlight the perspective of Prem Kumar, CEO of Humanly, thought leader in ethical AI, and past participant in the SHRM workplace Tech Accelerator Program. This week we are joined by Pallavi Sinha, who is the VP of growth at Humanly. Pallavi, over to you!

Thank you, Nell! I am currently the VP of growth at Humanly, a conversational AI start-up in the recruiting space. I am a demand generation marketer by background, and have a master's in business and marketing with 16+ years of experience spread across India, Australia and the US. I’ve worked with various Fortune 500s and global brands including Google, Mini Cooper, Nestlè, World Vision and Boost Juice. For the last 7 years, my focus has been supporting revenue growth in the B2B venture-backed, start-up space, including at employee performance management start-up, Reflektive. I'm passionate about data, marketing, building organizations, DEIB and spend my free time chasing wildlife globally and traveling.

I’m very excited to jump into this topic today.

Productivity Unleashed

Generative AI emerges as a powerful tool, filling gaps across the HR team needs, from recruitment planning and candidate engagement to offboarding, and everything in between. Unlike earlier AI applications such as recommendation systems, the latest systems can handle a diverse range of inputs and outputs. Connecting HR tasks to the business's "north star" involves understanding the performance equation: Engagement and Productivity.

• Engagement - is your team energized, empowered, and invested
• Productivity - is your team getting things done

The new performance equation is a dynamic interplay between team engagement and productivity, where both elements multiply each other, contributing to sustainable business outcomes.

AI Opportunities across the employee lifecycle

AI applications span the employee lifecycle, offering solutions that help automate various tasks, increasing efficiency and capacity for HR teams. These AI applications were once exclusive to organizations with significant internal machine learning talent. Now, various innovative technology solutions make them accessible, and their integration supports engagement and productivity, bolstering overall performance sustainably.

AI opportunities for engagement and productivity

• Recruitment planning - creation and adaptation of job listings, recommendations to optimize listings, sourcing based on skills and organizational goals
• Candidate search and application - sourcing, initial engagement, guided application process, qualification, initial engagement insights
• Screening - chatbot assisted screening, engagement focused insights, anti-bias tracking, company positioning and fit
• Interviewing - scheduling, suggested interview questions, note taking, summary of interviews, interview training and AI-generated insights
• Offer and onboarding - offer generation and negotiation support, personalized and guided onboarding, automation of onboarding logistics, background verifications, immigration, paperwork and device allocation automation
• Performance management - feedback generation, coaching nudges, training, tracking of career goals and opportunities and insights into performance metrics
• Learning and development - personalization and learning content delivery, internal skills mapping, recommendations for continuous learning, updates to digital persona profile and internal skills upon completion
• Team and growth - personalized rewards and recognition, insights for team collaboration, recommendations for projects and gigs and updates to digital persona profiles
• Engagement, listening, and wellbeing - Team pulse surveys, coaching triggers tied to actions, recommendations on benefits, and well being, sentiment analysis and employee listening
• Offboarding - personalized guidance during offboarding process, dynamic offboarding experiences, updated digital persona profile and automated offboarding logistics
• Alumni - personalized access to company programs, roles, and future opportunities, alumni relationship management and communication

Many of these applications weren’t possible or weren’t feasible for organizations without massive internal machine learning talent until recently. Today, there are a range of external vendors and routes to roll your solution for every opportunity area above. Combined and properly integrated, these solution sets can support both ends of the engagement and productivity spectrum, more fully supporting performance in a sustainable way.

Paradigm Shift: AI User Experience

The latest AI in human intelligence is marked by a revolutionary user experience. We can compare the user experience enabled by conversational and more contextually aware AI with that of legacy systems. This begins to highlight why AI is now mature enough to take human intelligence by storm.

Legacy (pre-determined) UX:

• Linear and hierarchical UX
• Requires navigation
• Higher cognitive load and tool knowledge requirements
• Complex and non-intuitive data taxonomies
• Static and inflexible user contexts

AI (just-in-time) UX:

• Precise surfacing of tools and interfaces (users “teleport” to the right place)
• Bypassing of complex navigation
• Micro-interactions at the right time for lower cognitive load
• Simplification of complex data taxonomies (more natural language)
• Dynamic user contexts
• Algorithmic personalization

This evolution ensures the delivery of tools and insights precisely when and where needed, transforming HR operations for increased agility and responsiveness. In turn, leading to significant improvements in the overall employee experience, fostering a more engaged and productive workforce.

Understanding your Data: Guardrails for Success

The promise of the next generation of human intelligence is being progressively realized. It comes with its own set of considerations and new requirements of understanding. As with many business functions, both leaders and contributors within human capital management are finding the need to educate themselves more deeply about how AI works, how it can be audited, and the metrics needed for performance assessment.

One of the key differences between legacy platforms and AI platforms is that traditional platforms generate data, whereas AI platforms synthesize and manipulate existing data.

Conversational AI and natural language processing can utilize data with much greater flexibility than traditional platforms.

Just as humans “are what they eat,” AI is only as good as its data quality, relevance, transparency, and explainability. Without robust understanding and processes to ensure standards are met, sleek new tools can perpetuate bias and lead to poor outcomes, including financial, legal, and regulatory risk.

Guardrails play a crucial role in managing AI risks. For instance, Humanly's product suite includes a conversational chatbot for engaging job applicants and video analytics for interviews, both requiring careful control over data sources and communication. Humanly’s ethical AI manifesto details some of the guardrails in place to effectively manage AI risk. Guardrails like this serve as tactical responses to real-world challenges, ensuring AI aligns with organizational values and mitigates risks effectively.

Now that the stage is set, let’s examine some core metrics and why they are important.

What is data quality, and why is it important?

Data quality refers to the accuracy, completeness, and reliability of data. High-quality data is crucial for AI systems because it ensures that AI's decisions, predictions, and actions are based on accurate and reliable information.

Prem: "Ensuring data quality is foundational for any AI-driven process. Without it, we risk basing important HR decisions on flawed information, leading to biased outcomes and undermining trust in AI systems."

What is data relevance, and why is it important?

Data relevance involves the pertinence and applicability of data to the specific context or problem being addressed. Relevant data ensures that AI systems focus on the right information to make informed decisions.

Prem: "Data relevance ensures our AI tools are aligned with the strategic objectives of HR and recruiting, filtering out the noise and focusing on what truly matters for decision-making and insights. Simply put, the more relevant the signal you’re feeding the AI, the more relevant its recommendations and performance."

What is transparency in AI, and why is it important?

Transparency in AI refers to the ability to understand and trace how AI systems make decisions. It's important because it builds trust among users and stakeholders by making AI processes clear and understandable.

Prem: "Transparency in AI allows us to demystify AI decisions, ensuring all stakeholders can trust and effectively audit AI outcomes, fostering accountability and confidence in AI applications. Particularly when we’re talking about engagement and performance, without confidence, you can’t reliably support these goals."

What is explainability, and why is it important?

Explainability is the extent to which the internal mechanisms of an AI system can be understood by humans. It is crucial for diagnosing and correcting biases, errors, or inefficiencies in AI systems.

Prem: "Explainability bridges the gap between AI decision-making and human understanding, enabling us to validate and improve AI models continuously while ensuring they align with ethical and operational standards. It’s essentially a subset of transparency that supports the ability to precisely describe how decisions are being made. "

Finally, guardrails for AI are something you’ve spoken about a lot. What do you mean by guardrails, and why are they important?

Prem: “Guardrails are the tactical response to any real-world curve balls that can make your AI a risk. For example, at Humanly, one feature set is a conversational chatbot for engaging job applicants, getting them through initial logistics, vetting them at scale, and answering questions from candidates. It’s a voice for your company and handles more conversations than any one human resources team could manually handle. In short, it represents our customer’s company. So you need to have control over what it can say, where it draws its data from, how it communicates, and so forth.

A second portion of the product provides video analytics around interviews like a copilot. Guardrails factor in here in a different sense. Our AI can rapidly draw from many ‘live’ and existing data sources, monitor the situation to determine both best practices and recommendations about candidates, and flag instances of hiring bias. In this sense, it’s a guardrail around real-world outcomes. You have companies that provide auditing and monitoring for regulations, which is really important for a strategy and routine health of systems. You also want tactical ‘in the moment’ responses in the form of guardrails.”

Culture addition at scale

One emerging paradigm shift we haven’t touched on, regarding AI in human intelligence, is toward culture addition, even in high velocity hiring scenarios.

AI insights that can enable culture addition include:

• Process-level monitoring: Are recruiters and interviewers weighting the right details about interviews? What are the most effective techniques for gaining a full understanding of candidates?
• Examples: interview and video analytics, personalized suggestions for interviewers, anti-bias measures integrated with ATS data and co-pilots for interviews
• Tactical considerations: Are we surfacing the right career opportunities to the right culture addition (candidate, current employee, or alumni)? How can we get job listings in front of the right audience for culture addition? How can we limit loss of culture addition candidates due to a lack of engagement or new forms of engagement?
• Examples: automation of job listings and optimization, chatbots that can engage candidates throughout the hiring journey

From anti-bias protections to culture-building

Research indicates that diverse and vibrant work cultures lead to increased engagement and productivity. The range of elements that can contribute to a diverse workforce is well established. But inputs are extremely wide-ranging and manual in-the-moment interview processes can struggle to support hiring candidates who don’t look like those who have previously worked at an organization.

One trend that’s beginning to emerge is anti-bias measures within automated hiring systems, coupled with AI advancements, bringing a normative change in hiring culture. As models improve and more context is provided, the focus shifts from risk minimization to widespread confidence and adoption.

Moreover, the success of these systems hinges on their transparency and trust among all stakeholders. Open communication about how AI tools make decisions, audits and feedback loops can demystify the technology and foster a culture of inclusion and continuous improvement. This elevates the candidate experience and empowers hiring teams to make more informed, unbiased decisions. Ultimately, by prioritizing culture addition and leveraging AI responsibly, organizations can build for sustainable growth with more cohesive, diverse, innovative, and resilient teams.

Emerging players and established systems

Over the past year, we've witnessed a seismic shift in the landscape of emerging technology—a departure from the conventional trajectory. Suddenly, foundational models that once seemed exclusive are now accessible to all. The scaffolding for new solutions has grown exponentially, especially for those equipped with machine learning talent. Additionally, we’ve left a longstanding growth-at-all-costs start and scale-up market.

Within the realm of emerging players in AI for human intelligence, two notable trends have emerged:

• Differentiation through Deep Tech Advancements:
  The unique ability to stand out by leveraging advanced deep tech, which was once prohibitively expensive for most players.

• Emphasis on Sustainable Growth:
  A notable shift towards sustainable growth, often through strategic partnerships and engagement in use cases subject to increasing regulation and scrutiny.

This dynamic landscape makes AI in human intelligence an exceptionally fertile ground for collaboration between enterprises and startups. This collaboration extends on the product level and from a thought leadership perspective.

In this unique juncture, the convergence of enterprise wisdom and startup innovation presents unprecedented opportunities. For the first time in recent memory, startups are liberated from the relentless pursuit of growth at any expense. Conversely, enterprises are compelled to embrace a more agile approach, incorporating diverse perspectives into their strategies. Together, they navigate the evolving landscape of AI in human intelligence, shaping its future collaboratively.

Throughout this period, SHRM has remained on the pulse of both camps, maintaining long-term thought leadership for enterprise human capital management leaders while simultaneously expanding the SHRM Workplace Tech Accelerator Program and other initiatives.

For more research and thought leadership on AI in human intelligence, follow Humanly’s journey at Humanly.io or on Linkedin.