Artificial intelligence-based marketing is a scam.
We understand that it is the new buzzword, but it is not necessary to associate it with everything.
Artificial intelligence is gaining more and more relevance in various areas of our daily lives, especially since the launch of ChatGPT. This technology is viewed in opposing ways: for some, it is a threat that displaces jobs and raises copyright issues, while others see it as an opportunity to drive humanity into an era of advancement and knowledge.
The capabilities of tools like Midjourney, language models, and intelligent algorithms have radically changed the technological landscape. However, this advancement has the potential to generate both benefits and risks, depending on how it is managed and regulated. It is not surprising that tech companies are incorporating artificial intelligence components into their new products, seeking to persuade consumers of the necessity to purchase their devices.
From new AI-driven laptops to motherboards that optimize CPU performance, the trend is clear. Companies are urging their marketing teams to find ways to include AI in their offerings, especially at events like CES or Computex, often without any real impact on the value they provide to users.
A point of concern arises with the launch of the latest generation of CPUs by companies like AMD, Intel, and Qualcomm. Although these new products are impressive and show significant improvements in performance and efficiency, the addition of AI processors does not always imply real value for the consumer. Neural Processing Units (NPUs) have been introduced to assist in certain functions, but their relevance to the overall performance of the CPU is almost negligible in the context of artificial intelligence tasks.
After a month of evaluating various laptops and processors focused on AI tasks, it became clear that the true parallel processing power is offered by graphics cards. For example, when running benchmarking tests, an Nvidia RTX 4080 Super achieves astonishingly higher numbers compared to the latest CPUs. The reality is that AI tasks require a level of parallel processing performance that only GPUs can effectively deliver, as evidenced by Elon Musk's recent installation of 100,000 Nvidia H100 GPUs in an AI training system.
Furthermore, most popular AI tools currently rely on cloud computing to function properly. Large language models like ChatGPT need such an amount of processing power and storage that it is unfeasible to run them on local machines. Even some Adobe applications require cloud processing to handle images.
Despite the limitations of running most AI applications at home, the emergence of new laptops and chips with this technology continues to be a constant. However, in many cases, this may seem like an exaggeration that does not provide useful solutions to the problems faced by consumers.
