[For the uninitiated, here's a rundown on BotSlop](https://en.wikipedia.org/wiki/AI_slop#:~:text=%22AI%20slop%22%2C%20often%20simply,connotation%20akin%20to%20%22spam%22.) It will matter a lot in 2026 and beyond. This is not news, per se. Reams of digital ink spills near daily on the [future of creative industries](https://www.bbc.com/future/article/20241018-ai-art-the-end-of-creativity-or-a-new-movement) in a world of AI. ### We believe much of the hype is - at present - misplaced. But that's just for now. ^fc3430 Think how everyone and everything was 'going green' in the '90s. Hype bubbles blow up until they pop. Which is not minimize the effects of [[LLMs]] and explosive growth in processing power. It's a big deal for everyone, but maybe not a singular revolution for anyone. Yet. Our core takes - so far - *vis a vis* what we see in the ad industry: - It's a debatable epistemological proposition whether LLMs can make "new knowledge". If you're really, really, alarmingly good at refining cliches... are you actually making something new? But then isn't [all great art theft?](https://creativityclasses.com/good-artists-copy-great-artists-steal/) That's for philosophers of technology to settle, not us. ^5d1ae8 - Once the fax machines and dinosaur bones are cleaned out of the offices of broadcast stations - or possibly even before - most ad placements will be programmatic. That is: you buy the audience, not the inventory, on a bidding system. Call it AI or not; that's just raw processing power finally overtaking the datasets that fuel ad placement. - There will probably still be bulk cost savings to be found, however, by buying entire platforms for reach within geographies. Which matters more in politics than the industry might want to admit. We get high on our own supply about "targeting". If the challenge is winning "bids" in [[The Digital Attention Economy]] and [[Quote on Intellectual Obsesity|simply getting people to watch]], why not target as many people as possible? - Because it doesn't make you sound smart in a pitch, for one... - [Automations in image editing are slashing the time it takes to make individual creative assets; but not the skill or judgment necessary to make creative that works](https://x.com/legalmenteLugo/status/1907899816643850277). - Logging footage, for example; will never be the same - and that's a good thing. Automation will yield better products, more efficient use of agency resources, and lower barriers to entry for skilled creatives or diffuse teams. - Copywriting will change but the [[Copywriting Rules for Political Scripts|need for finesse will remain]]. - Churning out quick variations on a theme is now as easy as pushing a button. - Asking AI to sample for tone ("How does Barack Obama talk about race?") is a fascinating, insightful exercise that can make good writers better. - [It might not be the time to get into voice acting](https://www.hume.ai/?gclid=CjwKCAjwuIbBBhBvEiwAsNypvSN7V5X1_vTbcgjbi6lGbuh5Y4YeWARNYnZLm55NVQSzE_LNBSm2qhoCuY8QAvD_BwE&campaignid=22291527240&adgroupid=&gad_source=1&gad_campaignid=22301914243&gbraid=0AAAAA-9v-glP3ewm7fzwN-bvDQLo3urlj). - The market for diversified *and* skillfully constructed content will *grow*. [[Every Image Matters (The Power of the Frame)|Being personal, trustworthy and credible to your audience will go from important to essential.]] - But AI reproductions might be useful to personalize otherwise authentic messages, or hammer frequency in nooks and crannies of the internet. - Good discussion on that [here](https://podcasts.apple.com/us/podcast/the-future-of-political-consulting-by-the-aapc/id1770372104) ##### For now, we think the more interesting question is how voters will react to the explosion of political information coming their way by campaigns, agencies, or activists who can use these tools -- just not very well. More - and more, and more and more - replicable content is coming. With no real controls on information. And algorithmic platforms will continue to reward salaciousness, [[Feelings, Facts, and Our Crisis of Truth|truth be damned]]. To quote [[Dana Gioia]] (not his brother [[Ted Gioia|Ted]]); if all creative culture is outsourced to the market, all we'll be left with is Tiktok and Pornhub. The internet feels headed that way. In the short term at least, we think that will make voters' eyes *even more attuned* to the replicative fakery so common in the political ad industry. (And of course [[Voters don't read|they don't read]]. So getting them to watch at all is the first and most critical step.) But remember: the bots aren't, by definition, going to make new ideas. They're going to recycle old ones with mind boggling consideration for repeating phenomena. Imagine a Political Botslop checklist: that is, if you had to rely - like a machine - on *only* a refined understanding of cliches and context clues to give your piece of creative any hope of a sensible structure - you'd probably settle on a checklist of inputs like this: ### The Botslop Checklist ^d3dbe4 > Candidate's mouth moving in slow motion ✅ > Supporters standing in a posed semi-circle: ✅ > Underscore reminiscent of a [Lawrence Welk retrospective](https://www.youtube.com/watch?v=CHnFO9g2vUg): ✅ > Voiceover toned and paced like a [1990s Folgers commercial](https://youtu.be/S7LXSQ85jpw?feature=shared&t=9): ✅ > Script jammed with sequential, poll tested nuggets absent visual context✅ > Other words that don't mean things ✅ #### Now go plug "make a political ad about XYZ" into your AI tool of choice. [[An Evolving Collection of Politicians Pointing at Jobs and Leadership|We suspect it will look familiar.]] (PS - If [[Every Image Matters (The Power of the Frame)|every image matters]], maybe don't work with agency that will outsource their choices to an algorithm.)