ChatGPT enhances creativity and problem-solving in ways that traditional search tools can’t match. Photo courtesy of Rice Business Wisdom

We all know ChatGPT has forever changed how we do business. It’s modified how we access information, compose content and analyze data. It’s revolutionized the future of work and education. And it has transformed the way we interact with technology.

Now, thanks to a recent paper by Jaeyeon (Jae) Chung (Rice Business), we also know it’s making us better problem solvers.

Key findings:

  • A recent study finds ChatGPT-generated ideas are deemed an average of 15% more creative than traditional methods.
  • ChatGPT enhances “incremental,” but not “radical,” innovation.
  • ChatGPT boosts creativity in tasks normally associated with human traits, like empathy-based challenges.

According to the study published in Nature Human Behavior by Chung and Byung Cheol Lee (University of Houston), ChatGPT enhances our problem-solving abilities, especially with everyday challenges. Whether coming up with gifts for your teenage niece or pondering what to do with an old tennis racquet, ChatGPT has a unique ability to generate creative ideas.

“Creative problem-solving often requires connecting different concepts in a cohesive way,” Chung says. “ChatGPT excels at this because it pulls from a vast range of data, enabling it to generate new combinations of ideas.”

Can ChatGPT Really Make Us More Creative?

Chung and Lee sought to answer a central question: Can ChatGPT help people think more creatively than traditional search engines? To answer this, they conducted five experiments.

Each experiment asked participants to generate ideas for solving challenges, such as how to repurpose household items. Depending on the experiment, participants were divided into one of two or three groups: one that used ChatGPT; one that used conventional web search tools (e.g., Google); and one that used no external tool at all. The resulting ideas were evaluated by both laypeople and business experts based on two critical aspects of creativity: originality and appropriateness (i.e., practicality).

In one standout experiment, participants were asked to come up with an idea for a dining table that doesn’t exist on the market. The ChatGPT group came up with suggestions like a “rotating table,” a “floating table” and even “a table that adjusts its height based on the dining experience.” According to both judges and experts, the ChatGPT group consistently delivered the most creative solutions.

On average, across all experiments, ideas generated with ChatGPT were rated 15% more creative than those produced by traditional methods. This was true even when tasks were specifically designed to require empathy or involved multiple constraints — tasks we typically assume humans might be better at performing.

However, Chung and Lee also found a caveat: While ChatGPT excels at generating ideas that are “incrementally” new — i.e., building on existing concepts — it struggles to produce “radically” new ideas that break from established patterns. “ChatGPT is an incredible tool for tweaking and improving existing ideas, but when it comes to disruptive innovation, humans still hold the upper hand,” Chung notes.

Charting the Next Steps in AI and Creativity

Chung and Lee’s paper opens the door to many exciting avenues for future study. For example, researchers could explore whether ChatGPT’s creative abilities extend to more complex, high-stakes problem-solving environments. Could AI be harnessed to develop groundbreaking solutions in fields like medicine, engineering or social policy? Understanding the nuances of the collaboration between humans and AI could shape the future of education, work and even (as many people fear) art.

For professionals in creative fields like product design or marketing, the study holds especially significant implications. The ability to rapidly generate fresh ideas can be a game-changer in industries where staying ahead of trends is vital. For now, take a second before you throw out that old tennis racquet. Ask ChatGPT for inspiration — you’ll be surprised at how many ideas it comes up with, and how quickly.

-----

This article originally appeared on Rice Business Wisdom. Based on research by Jaeyeon (Jae) Chung and Byung Cheol Lee (University of Houston). Lee and Chung, “An empirical investigation of the impact of ChatGPT on creativity.” Nature Human Behavior (2024): https://doi.org/10.1038/s41562-024-01953-1.


Here's what you should learn from social media influencers for your own business marketing. Photo via Getty Images

Key business marketing tips to take from top social media creators, according to these Houston researchers

houston voices

Influencer marketing is booming, with companies allocating 10 to 25 percent of their advertising budgets to influencer-led strategies. Between 2016 and 2020, the number of sponsored posts rose from 1.26 million to 6.12 million, and overall spending in the past few years has grown by billions.

When partnering with online ambassadors, brands certainly want a large influencer audience. However, audience size does not necessarily reflect the amount influencers are paid. Influencers with similar-sized audiences can be paid very different amounts.

That’s partly because brands also want an engaged influencer audience. An influencer may have many followers, but if those followers don’t actively interact with content, the influencer’s reach is limited. Engagement metrics like comments, shares and “likes” are often a more reliable indicator of impact than follower count alone.

The problem brands face — no matter who the influencer is — is that sponsored posts typically see a plunge in engagement, making it difficult to measure their success. Very little research examines this effect and how influencers can mitigate it.

In a new study, Rice Business professors Jae Chung and Ajay Kalra take up this issue, along with Stanford professor Yu Ding. According to the researchers, one way of boosting engagement overall, even on sponsored content where engagement often falls, is for influencers to increase audience perceptions of authenticity, perceived similarity, and interpersonal curiosity.

Even in a world full of filters and careful staging, authenticity is a key differentiator for leaders, businesses and personalities. One powerful way of appearing true to one’s own personality or character is to effectively share life stories. But social media influencers walk a fine line between presenting their authentic selves and monetizing their platforms.

To attract followers and content sponsors, influencers must curate the images they share, the words they say, and the timing and cadence of their posts. It’s a delicate dance between providing value through a genuine audience connection and aligning with brand interests.

Here are three simple but powerful ways that influencers can boost engagement by highlighting close relationships:

  • Post photos that include one or two close friends or family members.
  • Mention friends and family in the caption.
  • Use first-person language (e.g., “I,” “my” and “we”).

Referencing close social ties is an especially powerful way to boost engagement. According to Professor Chung, “Intimate social ties can make influencers seem more authentic and sponsored messaging seem less transactional.” This effect holds true even when controlling for variables like gender, frequency of posting, use of emojis and hashtags, and audience familiarity with the influencer.

The team analyzed over 55,000 Instagram posts from 763 top influencers during the second half of 2019. One of their most distinctive findings is that, in terms of boosting audience engagement, the ideal number of faces in a photo is three — the influencer plus two friends or family members. For an Instagram audience, this numerical face count proves a surprisingly effective metric for assessing the closeness of relationships.

Influencers can also seem more genuine to followers by referencing intimate social ties in their captions. Terms like “grandpa,” “bestie” and “soulmate” give followers access to an inner circle usually reserved for loved ones, making them feel more connected and invested in the influencer’s world and worldview.

In one experiment, study participants were shown a series of Instagram posts supposedly written by actor Jessica Alba. Testing the impact of language on the perception of close ties, the researchers wrote three different captions for the same image. One caption mentions Alba’s daughter (“Styling by my daughter. Isn’t this outfit cute?”). Another references a distant tie (“Styling by designer Kelmen. Isn’t this outfit cute?”). A third post provided a baseline by indicating no ties at all.

Study participants were asked to select which posts they liked most. The results supported the research hypothesis. Posts mentioning close relationships are significantly more likable than posts mentioning distant ties or no ties.

The team also examined the impact of expressing emotion on Instagram. Does sharing feelings — either positive or negative — help or hurt audience engagement? Using the Linguistic Inquiry and Word Count (LIWC) language processing program, the researchers categorized and analyzed the strength and valence of emotion-related words and emojis (e.g., “love,” “nice,” “frustrated,” “sad”).

What they found is surprising. Expressing emotion boosts audience engagement, perhaps because it bridges a perceived gap of celebrity between influencer and audience. But what’s interesting is that negative emotions are more powerful than positive ones. According to the researchers’ dataset, negative emotions are expressed only 9.08 percent of the time, while positive feelings are shared 36.03 percent of the time. So, one way of interpreting the finding is that the comparative rarity of negative feeling could take some readers by surprise, and thereby incite a stronger sense of authenticity.

Importantly, all of these findings regarding audience engagement most likely apply to platforms where a gray line exists between private and public life.

And, on this note, the researchers warn against the potential for oversharing and exploiting family and friends for the sake of monetizing content.

But the study shows how brands can strategically sponsor posts that incorporate close ties in photos, express emotion, or share anecdotes in first-person language.

By quantifying tactics to achieve a greater perception of authenticity, the research provides valuable guidance on how to cut through the noise on social media. One of the paths to a more engaged audience, it turns out, runs through an influencer’s inner circle.

------

This article originally ran on Rice Business Wisdom and was based on research from Jaeyeon (Jae) Chung, an assistant professor of marketing at Rice Business, Yu Ding an assistant professor of marketing at Stanford Graduate School of Business, and Ajay Kalra, the Herbert S. Autrey Professor of Marketing at Rice Business.

Ad Placement 300x100
Ad Placement 300x600

CultureMap Emails are Awesome

5 must-know fall application deadlines for Houston innovators

apply now

Editor's note: As fall reaches full swing, Houston's innovation scene is calling on the latest batch of founders and startups looking to make a difference. A number of accelerators have opened applications. Read below to see which might be a good fit for you or your venture. And take careful note of the deadlines. Please note: this article may be updated to include additional information and programs.

Did we miss an accelerator or competition accepting applications? Email innoeditor@innovationmap.com for editorial consideration.

Texas Life Science Forum

Deadline: Oct. 2

Details: Ventures can apply to present at the 15th annual Texas Life Science Forum, hosted by BioHouston and Rice Alliance. Participants will meet during office hours with venture capitalists, tech scouts, corporate venture groups and angel investors, and present their pitches in a public forum. Pitches take place on Nov. 10 and office hours are held Nov. 11. Find more information here.

Greentown Lab's Go Make 2027: Advanced Carbon Materials with ExxonMobil

Deadline: Oct. 9

Details: Greentown Labs is seeking applications from startups developing novel carbon-based technologies for its latest Go Make cohort in conjunction with ExxonMobil. The structured accelerator is designed to facilitate validation activities and explore potential long-term collaborations with Exxon, according to Greentown. Founders will have the opportunity to engage directly with industry leaders to test, validate and scale their carbon technologies in real commercial contexts. The program tentatively starts on Jan. 20, 2027 and concludes June 16, 2027. Find more information here.

Activate's U.S. Fellowship Cohort 2027

Deadline: Oct. 30

Details: Activate supports scientists at "the outset of their entrepreneurial journey." It partners with U.S.-based funders and research institutions to support its fellows in developing high-impact technology. Its fellows receive a living stipend, research and development funding, connections from Activate's robust network of mentors and access to a curriculum specific to the program for two years. Applicants must have a bachelor’s degree and 4-plus years of post-baccalaureate scientific research, engineering or technology development experience. Their work must be based in the physical or biological sciences or related engineering disciplines. Find more information here.

Rice Innovation Fellows

Deadline: Oct. 30

Details: The Liu Idea Lab for Innovation and Entrepreneurship (Lilie)'s Rice Innovation Fellows program supports Rice Ph.D. students and postdocs in turning their research into real-world ventures. Participants receive $10,000 in translational research funding, co-working space and personalized mentorship. Candidates from all Rice engineering and science-related disciplines are encouraged to apply. Find more information here.

The TMC's Accelerator for Cancer Therapeutics

Deadline: Oct. 30

Details: Texas-based ventures and researchers developing a cancer therapeutics project can apply to this accelerator funded by the Cancer Prevention and Research Institute of Texas. The nine-month program runs February-September 2027 and focuses on market research, FDA regulations, intellectual property, licensing, finance, fundraising, legal and other critical areas for cancer-related ventures. Participants will complete the program with at least one grant submission and have the option to pitch to investors, corporate partners, media and other influential guests. Find more information here.

Houston college joins inaugural workforce accelerator supported by Google

hands-on training

Houston City College (HCC) is one of 15 community colleges from around the country to be selected for the first-ever Workforce Futures Accelerator.

The three-year effort is supported by Google.org, the tech company’s philanthropic arm, and led by the Association of Community College Trustees (ACCT), a non-profit educational organization that represents over 500 community, junior, and technical colleges. The accelerator focuses on helping colleges embed virtual, employer-sponsored training opportunities into short-term workforce training programs, giving participants access to opportunities that they might otherwise receive through internships or other "work-based learning" stints.

"The Workforce Futures Accelerator reflects the Houston City College mission of offering a high-quality, affordable education for workforce training and career development," Pretta VanDible Stallworth, HCC trustee and chair-elect of the ACCT board of directors, said in a news release. "Advancing student success and creating pathways to opportunities ensures that our students are well equipped to succeed and build a secure future in today's economy.”

Through the accelerator, HCC is tasked with fusing online project-based learning opportunities with its workforce education programs. The idea is to give students hands-on experiences working on projects sponsored by employers, allowing them to gain real-world knowledge in the process.

HCC will select two workforce programs that meet the accelerator’s criteria and insert into them into coursework. In the second and third year of the accelerator, the selected colleges are expected to scale the programs by adding instructors and programs to develop a network of to support their continued implementation.

“Participation by HCC will strengthen how we provide students with career-connected learning experiences that complement their classroom education and align with the needs of employers,” HCC Chancellor Margaret Ford Fisher added in the news release. “We are focused on ‘future forward’ strategies to meet the present and future needs of our region’s businesses.”

Two other Texas colleges were chosen to participate in the accelerator: Lamar Institute of Technology in Beaumont and Grayson College in Denison.

The remaining cohort includes:

  • Bergen Community College in Paramus, New Jersey
  • Central Louisiana Community College in Alexandria, Louisiana
  • Clark State College in Springfield, Ohio
  • Great Basin College in Elko, Nevada
  • Heartland Community College in Normal, Illinois
  • Hudson County Community College in Jersey City, New Jersey
  • Manchester Community College in Manchester, New Hampshire
  • Mesa Community College in Mesa, Arizona
  • Mohave College in Kingman, Arizona
  • San Joaquin Delta Community College in Stockton, California
  • San Juan College in Farmington, New Mexico
  • West Virginia University Parkersburg in Parkersburg, West Virginia

New report ranks Texas among top 10 states where AI could disrupt jobs

AI Workforce

A new nationwide report examining where AI could "reshape" the most jobs has ranked Texas No. 9 among the most at-risk states for AI job disruption.

The new SmartAsset report compared all 50 states and the District of Columbia to calculate the estimated percent of the workforce employed in the 26 occupations with the highest AI exposure, as determined by June 2026 research by the Virginia Economic Information and Analytics Division.

The findings revealed that 500,000 Texas workers, or 3.55 percent of the total workforce, are employed in occupations with "high exposure to potential AI disruption."

This also places the Lone Star State as the 9th most at-risk state in the U.S. where AI exposure can lead to "declining hiring demand, wage pressure, task automation, and other forms of disruption."

"States with larger concentrations of highly exposed occupations could experience more pronounced labor-market changes, particularly in roles where core tasks are more vulnerable to AI-driven restructuring," the report's author wrote.

Texas' biggest cities, like Houston and Austin, are known for their thriving tech and business industries, and the study noted that many of the occupations within those sectors are the most at risk. The Virginia Economic Information and Analytics Division said the top five most AI-exposed occupations in the U.S. are: mathematicians, proofreaders, correspondence clerks, court reporters, and media and communication workers. Additionally, computer programmers, database administrators, web developers, telephone operators, and communications equipment operators round out the top 10 most at-risk positions.

These are the 16 remaining occupations most exposed to AI disruption, in order:

  • Data Entry Keyers
  • Statistical Assistants
  • Office Support Workers
  • Interpreters and Translators
  • Database Architects
  • Software Quality Assurance Analysts
  • Medical Transcriptionists
  • Software Developers
  • Writers and Authors
  • Payroll Clerks
  • Web Designers
  • Miscellaneous Computer Occupations
  • Insurance Claims Processors
  • Telemarketers
  • Computer Numerically Controlled Tool Programmers
  • Bookkeeping and Accounting Clerks

A separate SmartAsset report from April 2026 found about 20.5 percent of Texas workers use AI to do their jobs in some capacity. That trend will continue to shift further as employers and employees choose to adopt — or reject — AI implementation.

Across the U.S., Washington topped the list as the state with the highest concentration of AI-exposed jobs, with nearly 5.7 percent of the state's workforce employed in the 26 most at-risk positions. SmartAsset said Washington's high prevalence of technology companies is a significant factor that skyrocketed the state to the top of the list.

"Home to major technology companies including Microsoft, Amazon, T-Mobile and Expedia, the state has large numbers of computer programmers and software developers, two occupations with high exposure," the report said.

Meanwhile, Mississippi ranked No. 51 with the lowest concentration of AI-exposed jobs in the nation. About 22,500 workers in Mississippi, or 1.93 percent of its workforce, are at risk for AI disruption.

The top 10 states where AI could reshape the most jobs are:

  • No. 1 – Washington
  • No. 2 – Virginia
  • No. 3 – District of Columbia
  • No. 4 – California
  • No. 5 – Utah
  • No. 6 – Maryland
  • No. 7 – Colorado
  • No. 8 – New Hampshire
  • No. 9 – Texas
  • No. 10 – North Carolina
---

This article originally appeared on CultureMap.com.