Why and for what purpose
Posted: Tue Jan 21, 2025 9:21 am
According to McKinsey, in the foreseeable future, up to 70% of all calculations involving generative AI will be performed in the interests of certain commercial customers, in the B2C direction; while the share of B2B will remain about 30%. In the case of B2C, analysts name the following as the main six areas of application of generative AI:
software development - mexico whatsapp resource both programs for end customers and analytical applications for interpreting and analyzing code,
creation of creative content, primarily marketing, for a wide variety of platforms,
development of targeted applications for automated attraction of new clients and communication with existing ones on the first line,
scientific and engineering applied innovative developments (for pharmaceuticals, materials science, logistics, etc.),
the use of smart bots as secretaries-referents to compile short summaries of the most important information from legal papers, technical documents, recordings of long conversations, etc.,
more sophisticated versions of AI referents for analyzing huge amounts of data, including multimedia and unstructured data, capable of making non-trivial conclusions without prior prompting (for example, finding anomalies in MRI scans without using any specific medical information, simply by comparing the proposed samples with extensive databases of reference scans of healthy people).
Experts expect that the most profitable in terms of potential profits, but also the most resource-intensive, will be AI applications for innovative developments in various engineering and scientific fields. The second group in terms of profitability is formed by the creation of creative content, attracting new clients, and complex AI assistants. Finally, the least revenue will be generated by the use of AI for creating software (there is nothing surprising here - people will have to check the code proposed by the machine in any case; the cost of an error here, since we are talking about B2C, is extremely high - and therefore third-party costs are high) and the creation of simple secretaries-assistants. At the same time, the costs of training models for the last group are insignificant compared to the first, so the most intense competition is expected in this segment of the emerging AI market.
Needed most of all
The following estimate made by McKinsey allows us to estimate the scale of the term "sharp" in relation to the rise in demand for generative AI. If, by the end of 2024, the demand of customers (B2C and B2B together) for generative AI applications in computing power was estimated at 0.2 quintillion (0.2×10 30 ) FLOPs, floating point operations; then by 2030 this demand will grow at least 125 times, to 25.0 quintillion FLOPs.
software development - mexico whatsapp resource both programs for end customers and analytical applications for interpreting and analyzing code,
creation of creative content, primarily marketing, for a wide variety of platforms,
development of targeted applications for automated attraction of new clients and communication with existing ones on the first line,
scientific and engineering applied innovative developments (for pharmaceuticals, materials science, logistics, etc.),
the use of smart bots as secretaries-referents to compile short summaries of the most important information from legal papers, technical documents, recordings of long conversations, etc.,
more sophisticated versions of AI referents for analyzing huge amounts of data, including multimedia and unstructured data, capable of making non-trivial conclusions without prior prompting (for example, finding anomalies in MRI scans without using any specific medical information, simply by comparing the proposed samples with extensive databases of reference scans of healthy people).
Experts expect that the most profitable in terms of potential profits, but also the most resource-intensive, will be AI applications for innovative developments in various engineering and scientific fields. The second group in terms of profitability is formed by the creation of creative content, attracting new clients, and complex AI assistants. Finally, the least revenue will be generated by the use of AI for creating software (there is nothing surprising here - people will have to check the code proposed by the machine in any case; the cost of an error here, since we are talking about B2C, is extremely high - and therefore third-party costs are high) and the creation of simple secretaries-assistants. At the same time, the costs of training models for the last group are insignificant compared to the first, so the most intense competition is expected in this segment of the emerging AI market.
Needed most of all
The following estimate made by McKinsey allows us to estimate the scale of the term "sharp" in relation to the rise in demand for generative AI. If, by the end of 2024, the demand of customers (B2C and B2B together) for generative AI applications in computing power was estimated at 0.2 quintillion (0.2×10 30 ) FLOPs, floating point operations; then by 2030 this demand will grow at least 125 times, to 25.0 quintillion FLOPs.