The Rise of AI Startups: Transforming Industries and Creating Opportunities
The current boost in the field of artificial intelligence can only be compared to the beginning of the century, when the dot-com era gave birth to a huge number of companies in one way or another connected with the web. Not all of them succeeded.
Evgenii Chebotov, syndicator and co-founder of Prosto VC, explains how can a modern AI startups avoid repeating the mistakes of the past.
The explosive growth of interest in artificial intelligence in recent years is due to several reasons.
First, machine learning algorithms have improved dramatically, as well as the hardware base on which AI solutions run.
Second, enterprise systems have amassed a wealth of data on which to train AI.
And, perhaps most importantly, there are tasks that are virtually impossible to solve without the involvement of artificial intelligence.
In recent years, investments in AI have been growing rapidly. The leaders in investment are the United States with $47.4 billion, China with $13.4 billion, and the European Union and the United Kingdom with $11.04 billion. The main areas of investment in 2022 are medicine and healthcare ($6.1 billion), data processing and management ($13.4 billion).
Source: Statista
The rise in investment has caused an explosive growth in the number of AI startups, the last time this happened in the dot-com era. Not all of the startups from 20 years ago have succeeded.
And the startups of the 2020s face the question of how to avoid repeating the fate of thousands of companies from the beginning of the century.
“Where should I work then, what should I do?”
Conventionally, all AI startups can be divided into three categories. The first deals with AI hardware, from data centers to semiconductor components.
The second is the layer of software for working with data, these are tools for collecting, storing, processing and analyzing it, mainly in the B2B sphere. The third is the user layer, mobile applications for users.
So far we are witnessing the blossoming of projects of the “second layer”, but it is obvious that in the near future the funders of the “third layer” will become active and AI applications for smartphones will flood the market in huge quantities.
Working in the “second” and “third” layers, one should take into account the most important factor in the development of the AI market – OpenAI. This company literally exploded the market, reaching a user base of 100 million in just 2.5 months. Given that OpenAI is sharing its developments for free (and others will surely join it), the startup needs to realize that no one can predict what will be freely available tomorrow.
For example, a startup offers services using OpenAI’s open source code: improve recommendation algorithms of marketplaces, or take on automating typical business functions (writing requests, legal notices, sales department responses, etc.). Today, such services are in demand, but that can eventually change if OpenAI (or another company) puts out a publicly available solution that allows you to build these kinds of systems from a set of ready-made modules.
Another risk is potential market competition with large “traditional” companies. So far, on the wave of huge interest in AI all over the world this is not particularly noticeable. However, once the market stabilises, startups will quickly face a shortage of interest from customers who will prefer solutions from proven suppliers, especially since the giants will be outnumbered in both financial and human resources.
You can bet on the fact that it is easier for a giant to buy a startup than to compete with it, but it may not play out, especially if there will be a lot of people willing to sell themselves (and there will be a lot of them).
The AI is changing so rapidly today that it is too risky for a startup to create a full-fledged IT infrastructure and invest serious money in marketing. You need to look for specific niches in which you can use the open developments of the leaders, while avoiding competition with them.
There are various options for this. You can, for example, focus on the emerging business in the field of prompt engineering – the ability to write “correct” queries for GPT. This is a new direction that appeared literally from “nothing”, as previously the ability to write queries as the basis of a business model was not considered by anyone. Now many people want to get the right answers from GPT, and in order to get such an answer, it is necessary to know how to ask correctly.
This area is very likely to become a very important part of the world economy. And there are no clear leaders in it yet.
Another option for a relatively risk-free strategy is to work at the intersection of artificial intelligence and traditional business in some specific sector that does not interest large companies. For example, a lot of teachers today teach children the correct pronunciation, and so why not leave this responsibility to AI? It will not be difficult for AI to teach them to pronounce difficult sounds, for example, sibilants, to achieve correct pronunciation.
The use of AI in medicine is commonplace. Many companies of all sizes are working on the application of artificial intelligence in this field. However, the complexity of the human body gives a very wide field of activity, even if you do not go off the beaten track.
Similarly, the organization of a modern enterprise is quite complex, and production tasks (unlike “accounting” and record-keeping tasks), although typified in general, are very diverse when it comes to details. Accordingly, there are many niches for AI solutions that optimize specific processes.
Competing with OpenAI and other AI platform developers in such areas is unlikely; they are hardly expected to trade free creativity for delving into the details of physiological and production processes in the foreseeable future.
To summarize, we can assume that a startup that wants to succeed must be able to quickly create expert products based on publicly available solutions, if possible – at the intersection of industries, but sufficiently specialized to avoid direct competition with large companies.
And the only thing left to do is to beat similar startups that have chosen a similar niche to pour their efforts into. their energies
