Looking Ahead: 2025 Trend Predictions
As we head into 2025, we asked subject matter experts within the 18 Coffees community for their perspectives about where we go from here.
Cassie Marketos on Sustainability
Community composter and local climate advocate
Right now, even people who should be reasonably receptive to environmental concerns are not activated in the space. We have to change that! Get out there and get involved. Volunteer at your local community compost. (If you're in LA, I can help you with that!) Look up your local waterkeeper organization. Get trained as a naturalist, in order to better see and understand the natural world around you. These certifications are available in every state and cost a minimal amount. If you already have a job, start researching basic things you can do to advocate for a lower carbon footprint - from shifting what cloud storage systems you use to making adjustments to your supply chains. Do it now!!!!!!!
Cassie Marketos is the LA-based author of Compost This Book. Subscribe to her Substack, The Rot.
Jessyca Dudley on Philanthropy
Founder & CEO, Bold Ventures
Opportunities in Philanthropy for 2025
Despite these challenges, 2025 holds tremendous opportunities for transformation in the sector. There is a growing recognition of the power of trust-based philanthropy and community-led funding models to drive systemic change. More donors are embracing collaborative giving, participatory grantmaking, and redistributing wealth to historically underfunded communities. Technology and data also present opportunities to enhance transparency, measure impact more effectively, and scale solutions globally. At Bold Ventures we are optimistic about the potential for philanthropy to build stronger partnerships across sectors, amplifying efforts to address racial and economic disparities while fostering collective action.
Jessyca Dudley is a social sector leader and strategic advisor who specializes in guiding racial equity in philanthropy. Follow her work at Bold Ventures.
Andrew Paley on A.I.
Chief Science & Technology Officer and Cofounder, Storyline
The good news is that there are loads of opportunities here, but startups looking to raise on a landing page and a cool demo alone – with no sense of product-market fit, how to build into a user journey or workflow, or mechanisms to measure impact – are going to start to hit walls.
2. Ensuring teams that intend to leverage A.I. understand both the promise and the pitfalls: a lot of upskilling and awareness cultivation will need to happen as orgs of all shapes and sizes adopt A.I. strategies in their products and services.
3. Staying reliable at scale — tools and techniques to get the best of models while reducing the negative impacts of hallucination and other types of errors are going to be key as we build trust, maintain veracity, and check compliance boxes.
4. Fostering and maintaining the public perception of AI as a force for good, both within companies and within society at large. Our goal is to build things that make the world more understandable, tractable, useful, meaningful, inspiring – but there are plenty of opportunities for distraction, disinformation and economic displacement. Of course, we validate our claims to social good only by living up to them, and practitioners of artificial intelligence have an extraordinary responsibility to push in that direction.
5. Relatedly, the internet is going to continue to get a lot noisier and potentially less useful and trustworthy. We have a lot of work to do if we are going to rescue our online information spaces from the misinformation and noise, and as generative A.I. gets ever-better, the problem will only get worse in two ways: 1) misinformation in picture, video and audio form is going to become increasingly indistinguishable from the real thing, and 2) real information is going to become easier to doubt and dismiss as the result of A.I. The saying goes “democracy dies in darkness” but I think we might find out that it also falls ill in the blinding light, too.
What do you predict as the biggest opportunities for businesses in A.I. in 2025?
Many challenges are waiting to be reframed as opportunities: Reliability at scale will require grounding generative AI in verifiable, truthful, data-driven, and actionable information through increased integration of specialized models, analytics, knowledge graphs, databases and fact-checking mechanisms. The next wave of innovations will involve bringing new layers of trust, veracity and compliance to artificial intelligence now tempered by product and real-world requirements, especially as these tools and techniques further infiltrate highly regulated domains (finance, insurance, medicine, law, etc). We think about this a lot at Storyline, and we’ve built our generative runtime with truth maintenance at its core.
We’re moving out of the initial shock and awe era of AI — one where a thousand companies wanted to do something, anything, with language modeling and a million skunkworks projects kicked the tires on ideas that rarely made it past the demo stage — and into one where we’re refocusing on the original goal: to build products and services that solve problems in increasingly efficient, accessible, and meaningful ways. If the research hits a wall tomorrow — it won’t, not yet, but if it did — we have years worth of products and experiences that can be built with the set of tools we have on hand today. In general, those new possibilities come at two scales:
Going blue sky: A.I. is creating the space for new types of products and experiences that simply wouldn’t have been possible before. My own company, Storyline, is born of this reality – we’ve built a platform that would have been hard to imagine just a few years ago before advances in deep learning afforded us a new level of quality in generative A.I. There’s a lot of headroom now to think big, but it takes being conversant with the tools and techniques – knowing what they’re good for and where they’re likely to fail.
Going small (for big impacts in the aggregate): there are also a thousand small process improvements now made possible by different facets of A.I. and organizations have significant opportunities to take stock of the capabilities, and continue to standardize around those that reduce tedium, improve workflows, and lead to better results across the board. If you’re looking for a place to start, find the people in your organization that already leverage new tools and techniques and make them advocates and evangelists. Just make sure to have safeguards in place while validating the utility and reliability of new components.
I’m sure you’ve heard by now: Agents — models in ensemble frameworks that can work autonomously and proactively on various tasks that might include interacting with APIs and services as well as aligning and collaborating with other models. This will be a significant driver of innovation in 2025, though it will also be the source of new types of challenges, especially at scale.
Using internal data: Whether it’s domain-specific datasets or just the artifacts of business process themselves — emails, Slack messages, Jira tickets, customer support correspondence, etc etc — many companies are sitting on a treasure trove of data, and with privacy-preserving, partitioned, fine-tunable model deployments accessible through major cloud providers, more of those companies can begin to unlock and leverage its value.
Relatedly, many companies have demonstrated a renewed interest in getting their data houses in order over the past couple of years, spurred on by the possibility of leveraging that data in tangible ways through A.I. and analytics. That foundation is likely to drive rapid innovation in the year ahead. We spent 20+ years in the era of big data making the world machine readable and waiting for much of the promise to pay off, and advances in artificial intelligence are shaping up to be the back half of making that world of information tractable for and valuable to people again.
Andrew Paley is the Chief Science & Technology Officer and Cofounder at Storyline, where he works to turn data into personalized, A.I.-powered video for clients in financial services, retail, and insurance. Learn more at storyline.com.
Robin Kasner on Employee Engagement
Co-founder & Managing Partner, 18 Coffees
What do you predict as the biggest opportunities in employee engagement in 2025?
2025 will present opportunities for increasing employee engagement for companies that know how to use new A.I.-based technologies to enhance learning and development and meet the need for continued upskilling and reskilling. And the data continues to show that employees are asking for mental health and well-being support as a core element of fostering a supportive workplace culture going into what could be a stressful year.
Robin Kasner is a founding partner of 18 Coffees with 18 years of experience in communications, operations, and employee experience.