What is Obweb?

In 2016, 2 billion people will have cameras in their pockets.

These cameras are getting smarter.

The internet connects us to data. What if it could connect us to objects?

Obweb is a 2015 trendcast speculating the potential of computer intelligence + objects + humans with pocket cameras.

What is the internet with object recognition?

What behaviors does it create in communities and networks?

Initial Adoption

Popular culture resists new technology. How will culture adopt object recognition as an interface to information? What is the quickest way to establish dynamic object inventories?

Object cataloging through manual photo-captures feels cumbersome and slow. The initial application of object recognition has to promise enough practical and daily value for mainstream success. While surveillance feeds make people uncomfortable, its automatic nature makes it the easiest implementation.

In the near future, mainstream object recognition will debut most successfully in appliances like the refrigerator. Food objects challenge us with its diverse attributes—shape, expiration date, taste, uses, and more. Keeping track of this complexity can be off-loaded to a computer. Autonomous cameras in the fridge can dynamically and automatically relay the fridge's contents a computer.

A Fridge Tracker camera could allow:

  • an inventory searchable by query
  • awareness of expiration dates
  • notifications from unique food items
  • inventory-aware recipes and recommendations

What are the consequences of a widely adopted Fridge Tracker? Attitudes towards home surveillance might relax, or become inflamed if corporations take hold of this inventory information. In the example of a Fridge Tracker, a food inventory already reveals plenty about socioeconomic status, household size, and personal health.

An object inventory could then be understood as an individual's object-oriented profile—a dynamic set of objects tagged by ownership.

On a larger scale, these inventories become data sets representative of communities and cultural lifestyles. Once adopted, the broader concept of inventories could even shape object cultures (food, fashion, etc.) and the industries that manufacture them.

Inventories + Communities

How can we connect multiple inventories together in useful ways?

In the case of Fridge Tracker, connecting fridge contents creates simple shared inventories: for example, understanding that your neighbor has the ingredients you're missing. This service could also enable reasoning about a collection of shared inventories. In a shared food inventory, the act of borrowing one item could generate new recipes.

At a community scale, shared inventories create a sense of fluid ownership—for example, inventory sharing with local food banks can push specific food items into areas of need. This kind of culture could preemptively help people create meal portions that are more reasonable? for consumption and ask for a donation of whatever portions may be left over.

Eventually, inventories and services will extend beyond single categories to encapsulate many distinct types of objects. These categories might be defined in broad strokes ("food," for instance) or tagged with attributes at a granular level ("commonly deep-fried.")

The emergence of general inventories requires intelligent organization of object categories. This means distinguishing object behaviors and forming connections between similar attributes. On a network scale, general inventories collect into smart collections—rich databases of object knowledge.

To distinguish inventories from smart collections—where inventories know about category attributes and ownership, smart collections relate object behaviors across categories. In other words, smart collections refer to one or multiple inventories with connective intelligence.

Shared smart collections can function as marketplaces where individuals could buy, loan, or trade their objects. Easy access to community inventories enables a fluid and well-stocked sharing economy. When each object is stored and bundled with its usage, the marketplace even acts as an efficient and accessible workshop.

A user can search for how to build a table without any prior knowledge of carpentry. The shared inventory knows about the lumber, drills, and saws floating around the community. It presents the user with these objects, their rental fees, and their usage.

The Fridge Tracker expands an individual's food inventory into a shared food inventory, causing an explosion of possible recipes. General shared inventories expand an individual's entire material inventory, exploding the individual's potential across all material domains.

In the way that Google Search extends a user's factual knowledge, smart collections augment a user's creative ability.

Cultural Impact of Object Recognition

The sharing culture of smart collections will require a certain level of transparency in user inventories. These complete and public "object profiles" reflect lifestyle in a more transparent manner that isn't heavily curated like social media platforms.

Another thing to keep in mind is that the current companies that have the closest technology to object recognition are the technology giants: Facebook, Google, Amazon and Wolfram. How would they then use this technology to further their goals?

Could Facebook might use it to catapult social capital to new heights by creating mechanisms to distribute it through objects and people in the real world?

Could Amazon might solidify their status as the place to go for all your needs through using object recognition to more fluidly integrate shopping on their site into daily life? (See also: Amazon Firefly)

As of right now, the most developed object recognition programs are by Wolfram and Google. Wolfram Alpha's object-based search can already identify things like foods and animals with confidence. Wolfram software has consumed tens of millions of images, an amount "comparable to the number of distinct views of objects that humans get in their first couple of years of life" according to Stephen Wolfram. Meanwhile, Google's computer brain can talk about complicated images of objects with impressive natural language ability. Object recognition search products like Google Goggles have tried to funnel into mainstream use without much success.

Object recognition is still in utero, and is currently used for little more than a tech demo. There is great commercial potential for devices which understand objects and products; but the failure of Amazon Firefly and Google Glass shows that this new technology must be carefully guided for mass adoption. Instead, companies must create services which give compelling value to their customers. We think that the successful object recognition services will not be glorified barcode scanners, but humanist tools for manipulating, navigating, and understanding objects and products.

Who made this?