Do you — like me — wake up, read the news, and wonder in that Dorothy Parker way, “What fresh hell is this?” Or perhaps you are innately optimistic in that “Always Look on the Bright Side of Life” way – a dark and lugubrious optimism? I gather that back in 2014, this 1979 Life of Brian song was the most popular choice of a song to play in UK funerals.

It is this vein that I attempt to embrace a gloomy optimism as I muse on the state of publishing at a scholarly society. In many respects, this is a follow-up to a recent and excellent Scholarly Kitchen post by Ben Kaube and Steve Smith, “Societies 2030: The Community Advantage in an AI-First World“. Their approach was to look at a 2025 report by Rob Johnson and Sarah Greaves, concluding that:

“…society publishers are at a crossroads, facing declining revenues, rising costs, and pressure to consolidate. Their data is hard to argue with. Since 2019, over half of society publishers have seen journal revenues decline in absolute terms. Smaller portfolios and biomedical fields have been hit hardest. The trajectory is real.”

Despite this daunting picture, Ben and Steve were quite optimistic in their view of what a scholarly society may offer:

“Societies are different. Members build relationships with each other, reputations within the community, and professional identities tied to the field….”

They also argued (less convincingly, in my opinion) that the trust that scholarly communities feel towards their societies may make them more open to societies’ experimentation with AI tools. The trouble with this rather rosy picture is that distrust of AI has become deeply embedded in some parts of academia and indeed the wider public. As I was writing this post today I saw a recent fascinating and frightening article by Ece Yildrem in Gizmodo , entitled “People Hate AI Even More Than They Hate ICE, Poll Finds“. The article tells of a recent NBC survey of registered voters and discovered that 46% of respondents said they were negative about AI and only 26% were positively inclined towards AI.

“Even the Immigration and Customs Enforcement agency, which has been the subject of nationwide protests spanning months, was viewed more favorably, as was President Donald Trump, whose current net approval rating is -19.  In fact, the only topics viewed more negatively than AI were Iran and the Democratic Party.”

For a scholarly society to deploy AI tools for its membership, there needs to be trust and understanding in the promise and over-promise of such tools, along with a recognition that significant investment is needed, if those tools are to be trusted.

Which leads me to another interesting presentation I came across. This was in Wiley’s video series entitled “The Conversations, Episode 5: Navigating the future of scholarly societies“, a 30-minute long panel discussion that is worth a listen. The participants discuss what engagement with a scholarly society’s community really means: What is the value of membership? What levers are available to societies to achieve a sustainable, optimistic future?

Of course, the need for optimism is not a new phenomenon. On rereading one of my posts from 2014, I came across a sardonic musical reminder of reasons to be cheerful:

“Back in 1979, British New Wavers Ian Dury and the Blockheads produced the single, Reasons to be Cheerful (Part 3). The song was a celebration of sensory pleasures from music, comedy, food and beyond. Listening to this song recently, I was struck by its contrast with the tone of doom and gloom that exudes from the pores of publishers and librarians, and the need for a dose of optimism.”

Perhaps the world always seems to be in tatters around us. Perhaps there are positive messages we can grab onto.

While there are limited opportunities for driving new revenue, it seems clear we need to focus on community building, establishing member value and — to be completely blunt — on efficiencies and cost savings.

I work at a scholarly society — The American Mathematical Society (AMS) — the mission comes first. It is important we put the needs of our mathematical community out front. We emphasize developing accessibility tools for complex mathematics, recognize that we need to dig into issues of research integrity, promote a culture where author processing charges are not our norm, and focus on reasonable subscriptions and Green and Diamond models of openness.

Our community’s attitude toward AI is mixed: on the one hand, mathematics is driving the future of AI. On the other, mathematicians are skeptical of generative AI, not least because content quality may be diluted as AI-generated research ramps up and pollutes submissions. Authors want nothing to do with training generative models on their content without due attribution, compensation, or rigor of application.

So, where is my promised optimism? Where are my reasons to be cheerful, parts one, two, and three?

Reasons to be cheerful:

1. In recognizing the need to control costs, scholarly societies are being forced to look for innovative ways to be more efficient — and yes – to understand how AI efficiency tools can be used in the publishing workflow. No longer is it viable to do things as we have always done. We need to be maximally efficient – because someone else is likely out there using AI to be more efficient than us.
2. As a scholarly society, we stay close to our community and understand its needs. For example, we can reinforce a culture of research integrity as a member of our society — something beyond oneself — emphasizing publishing quality over quantity.
3. Storytelling becomes a priority. Translating our fields for those in and around our communities — and indeed, to a wider world of industry and policy makers — is now more essential than ever. A scholarly society can tell stories about people, ideas and in our case, the sheer beauty of math. This helps us define why being a member of our society is important to all mathematicians.

So, yes, I am optimistic, though darkly optimistic. Right now, while I see a focus on revenue generation as vital, it is efficiency and cost control that matter most.