A Better Internet: AI and Responsible Context

Copy to clipboard

Web 2.0 has been a battleground for your context

It’s no secret that the tech giants that underpin the internet have grown so powerful because of the data they have collected on us all as we skip our way through our online lives. Google tracks what we look for, where we shop, what we email and what we watch. Meta, TikTok, Snapchat, Reddit and et al. have intimate knowledge of the vast amount of content that we consume, comment on and share with our friends and foes. Even the current crop of AI giants are figuring out how to monetise your attention, by the time your read this OpenAI (ChatGPT) and Anthropic (Claude) may already be serving you adverts or embedding paid content in their responses when you chat with them.

This isn’t a groundbreaking insight from me, this is so well known at this point that you’re already losing interest in this article. Here’s a phrase that I coined to explain this succinctly: “If you're not paying for the product, you are the product.”. No I didn’t but it would be clever if I had.

Copy to clipboard

Where this had led us to now

The outcome of this battle for your context and engagement has had a number of negative effects which other experts and commentators are far better placed to discuss than I. However insofar as we’re talking about our online worlds in the context of well, context it’s useful to raise a couple of issues, aka ”externalities” if you’re boring enough to have studied economics like me. Firstly in the race to secure our valuable data the platforms have engaged in the most ingenious hacking of human psychology and neuroscience ever. They’ve built digital experiences which as so engaging on the fundamental brain-hardware level that our attention is hijacked for hours a day, more often than not in the full awareness that is it terrible for our own wellbeing.

All of this attention has been converted into a huge trove of data on each and every one of us - almost every bit of content we engage with online is tracked; what we did to get to that content, how we interacted with the content and what we did next. All of this information is refined by tech companies and turned into their best understanding of our personal context. If you think this sounds like a pretty poor way to get a true picture of an individual person then I would agree.

Below is a non-exhaustive summary of tools that have been developed for tracking what you do online, some of them (like cookies) can be used to help delivery better digital experiences, but most them amount to just lurking over your shoulder to get a better idea of who you are as a person and therefore who you are as customer.

We see how truly poor this method of building our personal context is every time we’re served irrelevant content (paid and organic) or we’re consistently presented with annoying stuff just because once upon a time we showed a vague passing interest in. How often are you following around the internet by that item that you left in your shopping cart somewhere or mentioned to a friend in chat?

The exploitation of our context is janky at best and a nefarious tool for manipulation at its worst. We are constantly being bombarded by nudges to go in directions that ultimately lead to nowhere or worse lead to a waste of happiness, energy and time.

Copy to clipboard

Let’s talk about context in context

I realise that I’ve just criticised the tech giants for hijacking your attention so I am going to be brief. Thankfully this isn’t a technical article on engineering so I can just plop out some briefly considered thoughts and see if anything sticks. I am also increasingly aware that I need to write stuff that doesn’t sound like AI so all the grammatical and syntax errors you find below have been carefully inserted and this isn’t just me writing long run-on sentences than run on and reflect a poorly considered point of view and a stream of consciousness after a few to many coffees.

For those of you that haven’t played with AI tools a great deal the idea of context might sound like a fancy technical word which engineers use to make AI appear abstruse (I learnt this word recently) and impossible to understand (this is what abstruse means). It really isn’t, context means exactly what you think it does - it is the relevant information that surrounds a particular domain that improves a person’s or machine’s understanding.

Think about it this way - A machine reads pixels and sees a dog catching a frisbee. A human sees a photo of their dead dog and feels a sense of loss for dear old Fido and nostalgia for their childhood home and garden. The latter is context - a machine cannot comprehend loss or nostalgia, but it can understand (with some follow-on questions) that the dog in question was yours and you used to live somewhere with a garden where you played frisbee with a large German Shepherd called Bash - AI also now knows you’re a geek.

Moving on from dead dogs: consider an AI-powered robot that you’ve asked to go food shopping, task context for it would be you telling it:

• What you want to eat for dinner
• When you want to eat

Seems obvious right, but you could imagine that mistakes could still be made if the robot didn’t already know relevant personal context like:

• That you’re allergic to dairy (check for ingredients for milk - which seems to be in everything)
• That your sad little kitchen only has a hob and you don’t own an oven or microwave

It would need to know certain additional personal context:

• You only have £25 left in your bank account
• Your supermarket loyalty card has active vouchers for half-priced orange-flavoured, chocolate-covered jelly discs (and you love Jaffa Cakes)

This is a stupid example but you get my point. You understand that the AI-robot has a legitimate need for context to best perform a specific outcome to your satisfaction. You understand why you need to share your personal context and you also understand why the ShopBot would need to do additional work on your behalf to check other environmental contexts that exist out there (e.g. opening hours, stock availability). Side note: AI models have fantastic environmental context after chewing up the entire internet, for example you do not need to explain that a supermarket is where you buy food, or what dairy-free means. AI models understand the concepts of cheaper, fresher, spicier etc. etc. This is why they are good at deriving intent and delivering outcomes.

Copy to clipboard

The old internet took your context and used it badly

In Web2.0 prior to going out shopping you might have searched for “Dairy free recipes for under £10 in hob-only kitchen”. The sneaky little internet spies shown in the table above might infer that you’re allergic to dairy, don’t own an oven and are broke. The internet would then start blabbing to pharmacies all about your allergies, then they would head over over to an appliance retailer and start shouting about your lack of oven and they might even decide to cross-reference what they infer about your financial situation with the credit data they have access to. The buy now pay later companies have heard that you’re strapped right now and are cleverly telling you can buy a compact air-fryer (worth it btw) and don’t need to worry about paying until a few months later when the interest alone will be eye-wateringly high.

Not only does this feel invasive, non-consensual and sort of shady it is also a really poor attempt at understanding you and your needs as a real human. You probably have your allergies under control and are fully aware that there’s no space for an oven in your little kitchen even if your landlord would let you install one.

Copy to clipboard

AI and consensual, relevant and responsible context

You might reasonably ask: in the era of widespread AI isn’t this problem only going to get worse? Companies who track you online now have this incredible power in their hands to crunch through the vast reams of data they’ve collected on you and target you even more aggressively. You would be right to ask this and yes this is certainly going to be a risk.

I would argue that there are glimmers of hope for a better internet and relationship with technology. This hope stems from the ownership and management of context. Let’s go back to our ShopBot example above. Let’s imagine that ShopBot is powered by two things: a food shopping-focused AI model and your detailed context. Let’s also imagine that ShopBot is driven by a single mission: shopping for your food. ShopBot won’t suggest news articles to you, it has no opinions on music and doesn’t care that you wear shorts in winter and you want the new foldable iPhone. It only cares about your financial situation insofar as it affects your food budget and it knows that anything that requires an oven is a non-starter.

In this world instead of your context existing as an inaccurate patchwork in the recommendation engines of tech companies, in resides within ShopBot’s single-minded and efficient AI brain. ShopBot is so effective at extracting semantic meaning and human intent from context that it knows after a week of just eating instant noodles you might need some vegetables and fibre. It might remind you that you’ve also committed to eating more lentils and want to reduce your meat-intake.

ShopBot is also ring-fenced so this context never goes anywhere without your consent and at the same time ShopBot will never attempt to use your context for anything besides buying food. You own (or probably rent) ShopBot and therefore you own your context. In this world you don’t mind sharing context about your allergies, kitchen, preferences and budget because you know it is going to be used responsibly in delivering the outcomes you desire. ShopBot is your trusted agent going out into the world with the right level of education about you to perform.

Your context is not going to be sold to the highest bidder who has paid for an advert which targets “dairy allergic medicine”, “small kitchen” and “broke” in their paid content advertising campaign. The actions which ShopBot performs in the world might send some signal to tracking companies, but ShopBot is focused on the task at hand - it cannot be distracted by Air Fryers (even sexy Ninja Air Fryers) or new iPhones. Maybe it reports back that Jaffa Cakes are on sale, you can make popcorn for cheap with kernels on the hob or that there is a new dairy-free ice cream brand on the market but your context is so well characterised that all of these facts are relevant and gratefully received by you.

An artist's impression of ShopBot

Copy to clipboard

Context and Network Effects

Maybe a friend wants to book a restaurant for your birthday surprise and maybe those friends care about what you might like to eat. With your permission ShopBot might be allowed to share certain specific content with your friend’s RestaurantOMatic (a french accented and obnoxious bot) to help them decide where to book for your birthday. You haven’t had a steak in weeks (because you’re broke) so you’re delighted to learn that your friends have booked FlatIron (great but you would have preferred Hawksmoor if they’re paying) - nonetheless they pickup the bill and you walk home licking your vanilla ice cream extremely content with the world - you’ve temporarily forgotten that you are allergic to dairy (you really don’t have your allergies under control). That’s a tomorrow problem.

Copy to clipboard

Wrapping up

Thank you for your attention and context (this blog has tracking that indicates that you’ve read this far in the article 🕵️‍♂️), here is a wrap up:

• A better internet future could be: moving context ownership from a poor whole-person patchwork owned by tech companies to a collection of highly detailed domain-specific contexts owned by the human
• Protecting context an ensuring that is is used consensually, responsibly and to achieve outcomes that are relevant to humans and the task at hand
• Creating rich context through the voluntary sharing of information and powerful extraction of semantic meaning and intent

Copy to clipboard

Plugging Hawkker

I’m currently building a version of RestaurantOMatic that we heard about above. It’s called Hawkker and it really does follow the principles in this article. Users provide rich context about what they like to eat, what they think of the places they’ve eaten at and why they are interested in eating at places they are yet to visit. This context is shared amongst food loving friends in the hope that we can all find better food to eat. You can join the waitlist at the link above