Image credit: Pexels
Four companies are using AI with human supervision and judgment to stay competitive.
The baseline for AI efficiency has moved for tech companies. It’s no longer simply about using AI; it’s about implementing it efficiently. Some companies are using AI to screen candidates for positions, restructure development teams, streamline workflows, and fill marketing channels with content.
Implementing AI was the first step, but continued business expansion in the current industry may depend on adopting AI strategically and keeping humans at the heart.
Screening Candidates
HiringBranch uses AI to assess soft skills like empathy via role-play. The tool uses proprietary linguistics and acoustics analysis to assess a candidate. The HiringBranch engine was built over six years and claims to have 98% accuracy in predicting on-the-job success.
Beth Thouin, the Chief Marketing Officer of HiringBranch, says, “…there are startups building AI interviewers and AI assessments based on whatever LLM is judging the skill set, and it’s a black box.”
She continues, “Our very big advantage right now is we have explainability. Every single one of our scores is defendable to a candidate, to an employer. It’s explainable because it’s proprietary.”
Transparency in scoring combined with a proprietary system seems to set HiringBranch apart.
Restructuring Development Teams
Bitsmith Studio restructured its development team from 8 to 5 developers using AI, enabling the company to refocus personnel without rehiring.
The AI generates code, effectively becoming a developer. The team can focus on higher-value tasks like problem-solving and quality assurance.
Muhammad Ali Abbas, CEO and founder of Bitsmiths Studio, comments, “The barrier of entry for junior developers has gone up because what they used to do can easily be done by AI.”
Junior roles previously involved generating code, which AI can do. They must now be able to judge code and have user-centric design skills.
Abbas continues, “Now they have to learn how to judge the code, what are the good coding practices, what are the bad coding practices, and not only write it, but also judge where it is written wrong and where it can be optimized.”
AI reshaped the development team’s focus and how Bitsmith Studios hires entry-level workers.
Streamlining Workflows
Formable, a contract management platform, uses AI to augment existing workflows, replacing high-volume repetitive tasks with AI, but the company does not allow it to create new workflows.
“If you’re trying to just find trivial ways to sprinkle AI into your work so you can say your organization is more efficient, but it’s not shifting the needle forward, it’s not super meaningful,” says Alex Shi, Founder of Formable. “If you have an employee and you want to delegate tasks to them, you have to first be good at doing that task yourself. It’s the same thing with AI.”
The Formable approach requires teams to master a task before using AI for it. Similar to Bitsmith’s approach, this shifts the skill set from performing the task to judging the quality of the AI’s work.
The goal of the company’s AI implementation is a positive, measurable outcome. In this case, lawyers create a playbook that is then fed to the AI.
Shi explains, “What our review feature does is read paragraphs in the contract, compare them to the playbook the lawyer has mandated. The AI basically does Control+F for you, and gives you suggestions.”
Understanding the tasks AI performs is foundational to implementing AI for Formable. Humans are built into the process as reviewers.
Fulfilling Marketing Content
52Launch assists businesses with design, prototyping, manufacturing, branding, and marketing. For the company, AI is a personal assistant that helps employees work faster and more productively, but not a replacement.
Peter Drakulich, Co-Founder of 52Launch, says, “The word ‘replacing employees with AI’ is wrong. The word is: how do I provide assistance to my employees that want to do things better, faster, and stay out of the mundane data?”
He continues, “It’s like those secretaries in Mad Men; paid a good wage to help make the person they were working for faster and better so they could stay at the top level of what they want to accomplish.”
Per company-provided data, 52Launch’s AI is about 80% accurate, but its accuracy can drift by about 15%.
Drakulich explains it as “the difference between getting to the North Pole or the South Pole at your starting point.”
As with several AI tools, humans correct it.
The AI produces up to 70 posts across 52Launch’s seven channels, which would not be possible without the automation AI performs.
An Augmentation Tool, Not a Replacement
Each of these companies implementing AI shares a central trait: they use artificial intelligence alongside humans.
From assessing soft skills to reviewing contracts, developing code to filling content calendars, these companies demonstrate an understanding of AI’s capabilities and limitations.
Human judgment is built into productivity within these workspaces and programs. Workers’ discipline, more than the technology itself, gives these industry professionals an advantage.
