Many sources show that the current AI voice market is around ten billion dollars and is going to grow at twenty-four percent CAGR. That is nuts. And I think, as we see stagnation happening in the AI coding market, voice is the next frontier.

In the last two years, coding giants like Cursor and all had their fun. We had Claude and all, but I think we are going to see deeper integration of AI voice tools everywhere. We have seen this happen with tools like Wispr Flow, Aqua, or even inbuilt tools by ChatGPT or Claude when we access their chat. The market is definitely there.

And another trend that we see is that everybody is working more. Corporations are expecting their employees to do more with the same time, and sometimes they might even ask that they work for more time because the competitors are also doing that. On the other hand, as a startup founder or whatever, you see your competition doing the same, so you might also have to do the same.

So the same human mind is being squeezed from all sides to produce more when it has not changed. So it will have to externalize a lot of things. We have to externalize our thoughts, externalize workflows. People have to soon figure out how to not become the bottleneck in the huge volume of work they are going to be expected to do.

And one of the fastest ways to get started is to reduce the friction between your mind and reality. Basically, if you have a thought or idea, if you see some information flying your way, you should get it down, note it down as soon as possible, because you are going to forget it most likely, and more barrage of information is going to come your way anyways.

So, having tools which just take stuff out of your brain and put it onto a computer, or capture stuff coming onto your computer into a more managed space, is necessary.

And currently, voice tools are what we see doing it the best. Instead of typing out, you are prompted. A lot of times, we are just saying it. Other parties’ meetings now get recorded and turned into AI notes.

So, voice-related tools will see a huge surge in the future, and the compute needed for them will also see a surge.

My thesis essentially is that we are going to have to work more. We are going to have more information to deal with, and voice is the way we can make this process faster than what it is now.

And since, in these two-three-year experiments with AI, we have understood that we need human supervision, humans are going to be needed in the loop. They are going to be fed a lot of information, and they are also going to need to process it and transmit it.

So reducing friction is going to be important.

AI increases productive capacity → organizations expect more output → organizations attempt more work
→ every human supervises more processes → more information flows through each human→ more decisions and instructions must leave → the human becomes the bottleneck → interfaces must become higher-bandwidth → voice becomes increasingly valuable.

AI will not immediately eliminate the human from most workflows. It will place the human above a larger number of workflows. That person will have to receive, understand, judge and transmit far more information than before. Voice will become an important interface for relieving that bottleneck.

And just like Shadow IT, we are seeing implementations of Shadow Voice AI. There are devices like Plaud and Neosapien, which are small variables or small devices, which can stay on during meetings and stuff and record the audio and give you transcriptions later so that in case you miss something from the meeting, you can just use it. So. Just like people want good health but can’t really make it practical to visit hospitals all the time, they just turn to supplements. We are seeing that the productivity pressures are making people go and opt for shadow AI solutions which are not really corporate authorized, but it is what it is. So we do see really strong yearning here.

Basically we are entering a new era of surplus, right? In past we had the also. Writing essentially started as accounting after we had started getting surplus for food from agriculture. Then we started to do computer stuff as we created more and more types of other surpluses, and AI is bringing just another one. But this time it is generative that it can. Create data on its own also. So the complexity is going to be a bit higher this time. Maybe the AI itself can absorb all that complexity, but a change is coming and it’s manageable. We have done this thousands of times. So we can look at the past how people dealt with things, how the legends did with things, how creative things emerged out of this. There is a lot of scope here. We as humanity can win.

Like for example, trade used to be concentrated in the region you are situated in. But with railway tracks and invention of telegraph, people from far away could communicate and trade at the same time. But the timing differences were there, right? So we came up with the invention of time zones. We always come up with something, and this time also we will do.

Basically humanity, We will figure out a way to create surplus every time and our old coordination systems and communication systems halt functioning in face of this surplus. And we always come up with something new to organize everything. And sooner or later we are going to have another wave of surplus and then we will have to deal with this problem again. But so be it. We are a species that repeatedly creates more than our existing systems can contain.

And when the systems become fragile, we start to upgrade them, either by hiring more people, improving them, or implementing new technology. And of course, you don’t wanna aspire to be hired because of this growth because it’s. Going to be shitty at start. Nor do you want to be consultant who will tell them how to get better at this. I think you want to be the one who owns the technology and sells it to them. Or in some way have some product that will let you ride this wave.