Artificial Intelligence dominates technology headlines. Depending on which article you read, AI is either about to transform society or replace it entirely. Yet for most businesses the value comes less from a dramatic overhaul and more from investment in practical AI applications.
The greatest benefit of modern AI is not that it thinks like a human. It is that it helps humans deal with complexity. Whether that’s understanding a lengthy business phone conversation, extracting meaning from thousands of emails, or making sense of large volumes of information. AI’s real strength lies in turning unstructured inputs into useful outputs.
That is where we’re seeing its value at Threads.
Back in the early 50s Alan Turing first coined the term “machine intelligence” to describe the behavior of a machine which a human cannot determine as non-human. An obvious example here is the computer. In order to achieve this, it is necessary to abstract away all the obviously non-human characteristics of the machine. This mostly means reducing the communication to something purely electronic – such as a telephone handset or computer terminal. So if I communicate with something at the other end of a telephone, say, and I believe that thing is a human, then according to Turing’s definition, that thing could be described as intelligent.
Over the past 70 or so years, AI has come to mean something different.
It is just as obvious that current robots are not humans and the ability to answer almost any question in seconds is not very human either. However, it is this ability that stokes up our fears, and the media loves it.
A short history of practical AI
But the concept of AI has been around for years as well as some of the very clever software which accepts and presents the information in a very human-friendly way. We call this Natural Language Processing (NLP). What has changed is the massive amount of publicly available information accessible within seconds. We call this information a Large Language Model (LLM) and it is created by continuously trawling the hundreds of billions of web pages on the Internet.
Search engines were the start of this, but typing in a search term and getting a long list of websites that each contain that search term is nowhere near as friendly as typing in a question in natural language and getting an natural language answer by analysing all those websites. This may look intelligent, but it isn’t and in this sense is no threat to civilisation.
Much is made of the fact that sometimes AI produces results that are plain wrong. However, these results are no more wrong than those obtained from a standard search engine and, in my experience, not more likely to be wrong than in asking a human expert.
Bringing structure to unstructured information
What makes AI genuinely useful is its ability to make sense of information that would otherwise be difficult, time-consuming or impractical to process.
You can provide it with a transcript, an email chain, a document, a collection of notes, or a mixture of all four. It can quickly identify themes, summarise content, highlight actions and uncover information that may otherwise remain hidden.
You can also ask any question and if that question has already been answered on a webpage, the engine will find it.
This ability to transform unstructured information into something useful is, in my view, the most intelligent aspect of AI today.
How we apply AI in our Threads Intelligent Message Hub
We use AI tools heavily in many different ways within our Threads Intelligent Message Hub, from business call transcription and call summarisation to making emails and conversations searchable across an organisation.
A phone call is one of the richest forms of business communication, yet it is also one of the most difficult to capture, share and search. Valuable information is often locked inside conversations, handwritten notes or individual recollections. Once a call has ended, finding a specific detail can be challenging and time-consuming.
Our call transcription technology converts conversations into searchable text, allowing businesses to locate information from phone calls as easily as they search their email inbox. AI-generated summaries then provide a concise overview of the discussion, highlighting key decisions, actions and outcomes without requiring someone to listen back to an entire recording.
However, the real value is not the transcription or even the AI itself. The value comes from making information accessible.
Practical AI applications beat the hype
Contrary to common perception, while the processes of NLP and AI can reasonably be described as “rocket science”, the use of those processes most certainly is not. The best analogy is a car automatic gearbox. Mechanically speaking, it is horrendously complicated, yet the effect of that complication is make the car so much easier to drive.
Its value lies in making the driving experience simpler, smoother and more accessible. AI is much the same.
The underlying technology may be highly sophisticated, but its purpose is not to impress people with complexity. Its purpose is to remove complexity from everyday tasks. For all the attention AI receives, its greatest contribution may be surprisingly simple: reducing friction.
Most people do not want artificial intelligence. They want faster answers, clearer information and less administrative work. AI succeeds when it helps achieve those goals without requiring users to understand the technology behind it.
The automatic gearbox transformed driving not because people became fascinated by gears, but because they no longer needed to think about them. AI is likely to follow a similar path. As the technology matures, it will become less visible, not more.
The future of AI is unlikely to be defined by machines replacing people. It will be defined by systems that quietly help people work more effectively by transforming complex, unstructured information into something useful.
At Threads, that is where we believe the greatest opportunities lie.
FAQ
One of the most practical uses of AI in business is helping people make sense of large amounts of information. AI can analyse emails, documents, phone call transcripts and other business communications to identify key themes, summarise content and make information easier to find and use.
AI uses technologies such as Natural Language Processing (NLP) to analyse text, speech and other forms of communication. It can identify patterns, extract important details and present information in a structured format, helping businesses save time and make better decisions.
Yes. AI call transcription converts phone conversations into searchable text, making it easier to review discussions, share information with colleagues and locate important details without listening back to entire recordings. It can also support compliance, record-keeping and knowledge sharing.
AI can help businesses organise and search communications more effectively. By analysing emails, call transcripts and other messages, AI can surface important information, create summaries and reduce the time employees spend searching for answers across different systems.
In most cases, AI is more likely to support people than replace them. Many of today’s most valuable AI applications focus on reducing repetitive tasks, improving access to information and helping employees work more efficiently, rather than replacing human expertise and decision-making.