Mastering AI Disaster: A Step-by-Step Guide to Company-Crippling Failure" (19 words)

Get ready to learn from two seasoned experts in AI failure, Greg and Hamill, as they share their 20-minute guide to ensuring the complete and utter destruction of your AI strategy.

1. Presenters, Greg and Hamill, will teach how to mess up AI strategy, embracing worst practices.
2. Greg is an executive leader with experience in AI strategies, ex-chief product officer at Pluralsight, and ex-executive leader at other companies.
3. Hamill is a machine learning engineer and independent consultant who has seen various ways AI strategies can fail.
4. The presenters have advised or interacted with representatives from numerous companies, aiming to create full-blown, company-crippling failures.
5. Begin by dividing and conquering your own company, creating impenetrable silos and incentivizing secrecy between teams.
6. Adhere to the "anti-value stick," focusing on unreasonable goals and promising customers the moon regarding AI capabilities.
7. Ignore details and focus on maximizing infrastructure costs, accepting a cascade of technical debt.
8. Build convoluted systems that only you, as an executive, can understand, ensuring job security.
9. When defining your strategy:
* Fake any diagnosis, highlighting random paragraphs from reports, especially the least understood ones.
* Be incredibly ambiguous and vague in your guiding policy.
* Create an AI-powered SEO tool, a generative art plugin, and an AI drone lunch delivery service.
* Announce all of this at a company All Hands meeting, using the word "disruptive" at least a dozen times.
10. Embrace perpetual beta, creating a massive backlog in GitHub and eroding people's willpower with extensive documents.
11. Use jargon strategically to hide jobs and confuse your organization, making sure nobody understands what you're saying.
12. Seed division in your organization by using technical language that alienates non-technical experts.
13. When mobilizing, randomly assign AI tasks to people with no relevant experience.
14. Outsource data review to offshore Q&A teams with little context about your business.
15. Launch untested, buggy AI chatbots directly to customers, disregarding quality assurance.
16. Focus on tools rather than processes, throwing tools at problems without analyzing or understanding them.
17. Blindly trust off-the-shelf evaluation metrics, ignoring customization for your business needs.
18. Avoid looking at data and let tools handle it, trusting AI outputs without verification.
19. Trust your gut and feelings when making million-dollar decisions, substituting data with intuition.
20. Ensure engineers are seen as coding wizards who can handle everything, even if they lack customer interaction.
21. Put data in complex systems only engineers can access, preventing domain experts from seeing it.
22. Insist on buying custom data analysis platforms that require a team of PhDs to operate and understand.
23. For real advice, visit AI-Exec or look for their O'Reilly book coming out February 27th.
24. The presenters are eager to help you on your journey, offering Q&A sessions after the presentation.

Source: AI Engineer via YouTube

❓ What do you think? What is one crucial step that can be taken to ensure the absolute failure of an AI strategy, as described in this presentation? Feel free to share your thoughts in the comments!