Why CanDo?

Now you can do anything 10x faster, cheaper, and easier without sacrificing quality, all while enjoying complete flexibility and having more fun. CanDo fulfills the promise of Artificial General Intelligence (AGI) like no other solution due to its proprietary architecture that has been decades in development.

 

CanDo's many benefits are listed in detail below, and contrasted with:

• Large Language Models (LLMs) like ChatGPT, Claude, etc.
• Virtual Assistants (VAs) that use Knowledge Graphs like Siri, Google Assist, etc.
• Bespoke Applications (BAs) of which there are millions.

CanDo allows you to mix visual interaction with the power of Language

You can seamlessly interact both visually and via language (written and speech), even combining them together, to create an extremely intuitive experience.

LLMs and VAs allow only textual (or voice) interaction, which then links to a separate integrated app (from a limited number of apps with limited functionality) for the visual part.

CanDo's visual interface dynamically adapts

The visual interface dynamically adapts how it appears and functions to provide the optimal experience for what you are wanting to do.

LLMs, VAs and BAs do not dynamically adapt.

CanDo is the first true No-Code platform

No need to ever get software developers to create a bespoke app, due to custom and complex requirements.

LLMs just generate basic code for developers, who then still need to integrate that code and check it for bugs. VAs require developers to integrate each concept and piece of functionality.

CanDo is an all-in-one solution

You can perform all the steps necessary to achieve your goal from within the CanDo platform. No need to use multiple apps to get the job done.

LLMs and VAs are often only useful for part of a task, leaving the rest for you to do via other apps, requiring significant additional cognitive load and wasted time.

CanDo doesn't need prompt engineering

You don't need to learn to think like the AI. Just express yourself naturally.

LLMs require prompt engineering, a skill that involves reason and discernment far beyond the norm where language is concerned. Prompt engineering is costly, time consuming, and highly specialised.

CanDo is extraordinarily flexible

You can fine-tune any details to fulfil your requirements, no matter how complex.

LLMs have a limited context size in which to specify requirements and don't perform well when faced with many or complex requirements. VAs depend on the level of flexibility of each integration, and this is usually quite limited and usually needs a level of knowledge of what can be done. BAs are only as flexible as the configurability developed for that application.

CanDo can perform any digital tasks

CanDo can perform any digital parts of any tasks and guides you through the rest. To interact with a 3rd party system, all it needs is access to an API.

LLMs can't be relied upon to perform tasks and don't oversee the coordination of the different parts of a task or larger goal, including parts that need human involvement. VAs are only capable of fairly simple tasks, and require a suitable app to be installed.

CanDo provides fine-grain control of what knowledge is available to it

You can select knowledge sources, assign a level of trust to each, and determine an overall order of priority (in case of contradictions). You are in command and can even tell CanDo to ignore specific knowledge within a specific source on a case-by-case basis.

LLMs have very limited and imprecise control over the data accumulated and the knowledge generated. A foundation or external model needs to be selected for each individual use-case (requiring experimentation to see which fits best), which then must be augmented with a custom external source. Additionally, LLMs use randomness to give the illusion of a human-like response.

CanDo's reasoning and logic are completely transparent

You can inspect the full chain of reasoning that leads to any decision.

LLMs use a limited and fallible linear chain-of-reasoning technique. And effectively how decisions are made are an impenetrable black box.

CanDo involves you in key planning decisions, then automatically takes care of the rest

You can bring your creativity, intuition and additional knowledge to bear in order to build the most appropriate plans and achieve the best outcomes.

LLMs take the reverse approach, putting the Machine before the Human. They don't engage you in important decisions. They give you the final response, which may be completely off the mark or significantly biased by the training data.

CanDo shows many relevant possibilities

You can select from a variety of the best possibilities, including superior alternatives that human experts may not be aware of (or can choose something altogether different).

LLM's produce responses that are an average of the training data. Regression to the Mean, a staple of AI, Machine Learning, and Conditional Probability, dumbs us down by suggesting the path already trodden by the majority. True creativity is impossible when the only data to draw upon comes from the past.

CanDo provides expert guidance throughout, including the pros and cons of each choice when presenting a decision

You can make informed decisions, without needing to possess any expertise.

LLM's often require you to think of the right questions to ask, putting a burden on you.

CanDo remembers every detail of prior interactions, including your preferences and goals

You can pick up where you left off, and never need to repeat yourself. And for things you don't want remembered, just say which areas or specifics are off-limits.

LLMs only remember what is in the current conversation context. VAs and BAs normally don't have any such memory.

CanDo adapts to your way of working

You can play to your strengths and natural ways of working.

LLMs, VAs and BAs do not adapt to your way of working.

CanDo tailors and optimises each experience according to the need

You can see just the relevant information and actions that are needed for the unique task at hand.

LLMs, VAs and BAs do not tailor the experience for you.

CanDo is your own personal self-development guru

You can receive expert teaching, mentoring, coaching, and training on any topic.

LLMs and VAs can't demonstrate the necessary initiative or expertly lead you through such processes.

CanDo doesn't need to be trained

You can simply tell it information (e.g. facts, rules, processes, etc), and it will immediately use them correctly and remember them.

LLMs require training that is time and resource intensive.

CanDo is efficient to run

You can deploy the platform with minimal resources.

LLMs boil the ocean to both train and run. The energy used to train a foundation model is like burning down a small forest, and the financial cost is massive.

CanDo provides fully optimised plans and implementations in an instant

You can leave the time intensive and difficult task of creating fully functional and optimised resources (e.g. apps, media, etc.) that match your requirements perfectly.

LLMs and VAs don't really implement. CanDo is able to optimise far better than BAs that are built by developers because it doesn't need to consider maintainability, modularity, and other factors that become non-existent.

CanDo is reliably accurate

You can count on the right actions being taken, which is crucial, as many actions can have significant consequences.

LLMs hallucinate, often giving out incorrect information and even fabricating entire stories. This is no small matter in an age of Post-Truth and widespread obfuscation. Determining the truth is increasingly difficult in today's world, and LLMs add to the problem. Even more importantly, they can't be trusted to perform tasks correctly, especially those involving real-world effects or having complex requirements.

CanDo is easy to keep within safe bounds of operation

You can set limits on what actions it can perform within various contexts (and CanDo already provides with a rigorous set of bounds out of the box).

LLMs are by nature quite imprecise in setting limits.

CanDo provides controlled self-learning

You can review and approve anything it learns, before it can make use of it.

LLMs are not able to offer this because the learnings are only implicit.

CanDo prevents collection of sensitive, private & protected data

You can easily prevent the ingestion of such data or keep it hidden to those without permission.

LLMs are challenged with preventing it, due to the way they work, and once in the system, can't easily limit who has access to it.