26 Oct 2024
Doc Searls writes, in Personal Agentic AI,
Wouldn’t it be good for corporate AI agents to have customer hands to shake that are also equipped with agentic AI? Wouldn’t those customers be better than ones whose agency is merely human, and limited to only what corporate AI agents allow?
The obvious answer for business decision-makers today is: lol, no, a locked-in customer is worth more. If, as a person who likes to watch TV, you had an AI agent, then the agent could keep track of sports seasons and the availability of movies and TV shows, and turn your streaming subscriptions on and off. In the streaming business, like many others, the management consensus is to make things as hard and manual as possible on the customer side, and save the automation for the company side. Just keeping up with watching a National Football League team is hard…even for someone who is ON the team. Automation asymmetry, where the seller gets to reduce service costs while the customer has to do more and more manual work, is seen as a big win by the decision-makers on the high-automation side.
Big company decision-makers don’t want to let smaller companies have
their own agentic
tools, either. Getting a DMCA
Exemption to let McDonald’s franchisees fix their ice cream machines
was a big
deal that required a lengthy process with the US Copyright Office.
Many other small businesses are locked in to the manual, low-information
side of a business relationship with a larger one. (Web advertising is
another example. Google shoots at everyone’s feet, and agencies,
smaller
firms, and browser
extension developers dance.)Google employees and
shareholders would be better off if it were split into two
companies that could focus on useful projects for independent
customers who had real choices.
The first wave of user reactions to AI is happening, and it’s adversarial. Artists on sites like DeviantArt went first, and now Reddit users are deliberately posting fake answers to feed Google’s AI. On the shopping side, avoiding the output of AI and made-for-AI deceptive crap is becoming a must-have mainstream skill, as covered in How to find helpful content in a sea of made-for-Google BS and How Apple and Microsoft’s trusted brands are being used to scam you. As Baldur Bjarnason writes,
The public has for a while now switched to usingAIas a negative—using the termartificialmuch as you do withartificial flavouringorthat smile’s artificial. It’s insincere creativity or deceptive intelligence.
Other news is even worse. In today’s global conflict between evil oligarchs and everyone else, AI is firmly aligned with the evil oligarch side.
Google, Microsoft, and Perplexity promote scientific racism in AI search results
Chatbot that caused teen’s suicide is now more dangerous for kids, lawsuit says
Thousands of creatives sign petition against AI data scraping
Authors who release under Creative Commons licenses are disagreeing with the CC organization about whether AI training is fair use: fair use alignment chart
The AI Boom Could Use a Shocking Amount of Electricity, and—Baldur Bjarnason again—Your use of AI is directly harming the environment I live in.
But today’s Big AI situation won’t last. Small-scale and underground AI has sustainable advantages over the huge but money-losing contenders. And it sounds like Doc is already thinking post-bubble.
Adversarial now, but what about later?
So how do we get from the AI adversarial situation we have now to the win-win that Doc is looking for? Part of the answer will be resolving the legal issues. Today’s Napster-like free-for-all environment won’t persist, so eventually we will have an AI scene in which companies that want to use your work for training have to get permission and disclose provenance.
The other part of the path from today’s situation—where big companies have AI that enables scam culture and chickenization while individuals and small companies are stuck rowing through funnels and pipelines—is personal, aligned AI that balances automation asymmetries. Whether it’s solving CAPTCHAs, getting data in hard-to-parse formats, or other awkward mazes, automation asymmetries mean that as a customer, you technically have more optionality than you practically have time to use. But AI has a lot more time. If a company gives you user experience grief, with the right tools you can get back to where you would have been if they had applied less obfuscation in the first place. (icymi: Video scraping: extracting JSON data from a 35 second screen capture for less than 1/10th of a cent Not a deliberate obfuscation example, but an approach that can be applied.)
So we’re going to see something like this AI cartoon by Tom Fishburne (thanks to Doc for the link) for privacy labour. Companies are already getting expensive software-as-a-service to make privacy tasks harder for the customers, which means that customers are going to get AI services to make it easier. Eventually some companies will notice the extra layers, pay attention to the research, and get rid of the excess grief on their end so you can stop running de-obfuscation on your end. That will make it work better for everyone. (GPC all the things! Data Rights Protocol)
The biggest win from personal AI will, strangely enough, be in
de-personalizing your personal information environment. By
doing the privacy labour for you, the agentic
AI will limit your
addressability
and reduce personalization
risks. The risks to me from buying the less suitable of two legit
brands are much lower than the risk of getting stuck with some awful
crap that was personalized to me and not picked up on by norms
enforcers like Consumer Reports. Getting more of my privacy labour
done for me will not just help me personally do better
#mindfulConsumption, but also increase the rewards for
win-win moves by sellers. Personalization might
be nifty, but filtering out crap and rip-offs is a bigger immediate win:
Sunday Internet
optimism Doc writes, When you limit what customers can
bring to markets, you limit what can happen in those markets.
As far
as I can tell, the real promise for agentic AI isn’t just in enabling
existing processes or making them more efficient. It’s in establishing a
credible deterrent to enshittification—if you’re trying to rip me off,
don’t talk to me, talk to my bot army.
For just a minute, put yourself in the shoes of a product manager with a proposal for some legit project that they’re trying to get approved. If that proposal is up against a quick win for the company, like one based on creepy surveillance, it’s going to lose. But if the customers have the automation power to lower the ROI from creepy growth hacking, the legit project has a chance. And that pushes up the long-term value of the entire company. An individual locked-in customer is more valuable to the brand than an individual independent customer, but a brand with independent customers is more valuable than a brand with an equal number of locked-in customers.
Anyway, hope to see you at VRM Day.
Bonus links
Space is Dead. Why Do We Keep Writing About It?