Your AI Survival Guide • Sol Rashidi & Joe Reis • GOTO 2024
Interesting. release "Goto Book Club" about artificial intelligence, in which Joe Reis interviews Sol Rashidi, author of the bookYour AI Survival Guide" And that's what this interview is all about. Sol talks about its experience with artificial intelligence: helped launch IBM Watson in the future. 2011He was Chief Data and AI Officer at Sony Music, Chief Data and Analytics Officer at Merck Pharmaceutical. She has been working in large corporations, but 8 She became an ex-C-suite executive. AI is now democratized and it is available not only to large companies, which is good, but because of this there is a universal hype. The book is devoted to real problems and failures associated with artificial intelligence, and how to avoid them. The choice of AI strategy in the company should depend on the maturity of the organization and its willingness to take risks.
- If the organization is conservative (many corporations)It is important to make sure that the strategy is focused on the inner world, not the outer. The bottom line is that we can let internal customers down when we try to implement AI, but chances are we’re not ready to experiment on external customers. At the same time, you need to keep big goals in mind, but start small and train muscles. (It is necessary to pick low-hanging fruits and show that the strategy works.) Saul is funny about the fact that when choosing an AI solution provider, you do not need to listen to marketing blatant salespeople, but you need to communicate with engineers and solution architects. (which are usually provided by providers) This allows you to understand how the solution really covers your needs and it can really be dragged into the market. According to Sol, open source is suitable for small companies or PoC, but managed services are more common in enterprises. This is partly due to the fact that enterprises generally like boxes, since cool techies are not ready to go to a conventional corporation, so corporations lack their own competencies. Joe asked what the future holds for AI, and Sol predicted three things. In the future, companies will begin to consider the cost of using managed services when assessing the ROI of AI innovations. This cost should be taken into account when implementing solutions. Now everyone is implementing AI, but not always bring the economy down.
- Now huge energy resources are spent on AI, both for training and for inference. This doesn’t go well with the ESG agenda and sustainability. Next articleTrends on AI and ESG will try to bring together Now the hype is unfolding around LLMs, which show state-of-the-art results in many areas. A lot of companies make foundational models, other fine-tunes make them for their needs. Saul suggests that there will be some managed middle-layer, which will allow corporations not to descend to the level of LLM. At the end, Joe asks Sol about how developers can improve their skills in AI, and Saul says that you should start with using artificial intelligence in your life and in solving your own problems. This can be useful for training. Saul suggests exploring and experimenting with various apps and players in the field. She also offers AI courses from top universities that are available on Coursera, Edx, Linkedin.
#DataScience #ML #AI #Data #PopularScience #Engineering