Les meilleures idées business de 2026 (Update ideas.xyz)

This video provides an update on ID, specifically introducing IDES V2 and discussing its evolution, challenges, and future direction. The V1 version of IDES had good ideas but lacked in execution and aesthetics. V2, which has been in use for several weeks, addresses these ergonomic and stylistic issues. It features a gauge, tags, various charts (which were also present in V1), and a reorganized opportunity matrix and market demand matrix. The speaker particularly likes the value scale section in V2. However, the main frustration with V2 stems from the gap between good business ideas and their final presentation. The speaker, with 10 years of experience in the business game and access to data, believes the initial ideas are strong. The problem lies in the human element and processes involved in translating these ideas into a polished output. The speaker admits a personal aversion to management, stating that working with people often leads to suboptimal results. There are humans, processes, air (presumably meaning inefficiencies or delays), and various tools for sourcing, organizing, and aggregating data. Yet, the final output is often mediocre, and sometimes, the core idea is even lost or misunderstood in the process. This issue highlights a significant problem in current AI applications: the combination of humans and AI often underperforms both AI alone and humans alone. A study, about 1-1.5 years old and from the medical field, compared AI alone, humans alone (with traditional analytical tools), and a human-AI combination. It found that AI alone performs increasingly well, even surpassing humans in medical diagnostics in many cases. Humans alone perform adequately. The surprising finding was that the human-AI combination performed the worst. This could be due to humans ceasing to think, disrupting the machine, poor communication between human and machine, or either party deferring responsibility. The speaker notes that AI tools like GPT can also be "lazy" and deliver subpar work at times. The current IDES system is described as a human-machine combination, which, based on the study, is suboptimal. The speaker plans to eliminate the human element in the process, aiming for an AI-only system. This decision is also influenced by development challenges. While "no-code" or "low-code" solutions have advanced significantly, the platform was initially built with a developer in late 2025 (presumably referring to a future date or a typo for 2015/2005) because no-code options weren't as robust then. Today, the speaker wants to add new sections and features, but the current lead developer is occupied with other major infrastructure work for a separate project (rich.4), which cannot be no-coded. Consequently, the speaker will rebuild the entire platform alone, without human developers, to create a purely AI-driven system. The vision for the future is an AI-centric business where the user (with AI tools and agents) generates revenue, keeping 100% of it (minus taxes) without expenses for media buying agencies, designers, or developers. The only human role potentially retained would be for customer support, as current AI support is deemed "bof" (mediocre) and human support is relatively inexpensive. The core rationale is that the human-machine combination underperforms, while AI alone and humans alone perform better. The speaker's role will be to provide creative input, and the machine will handle the rest. ID already possesses a vast backend database, filled with data from numerous business podcasts, strategies, levers, and "secrets," as well as information from Crunchbase, Product Hunt, and other sources. The goal is for the speaker to interact directly with AI agents that access this backend data to produce idea sheets. This is expected to lead to a higher quantity and quality of ideas and strategies. Regarding pricing, nearly all clients currently opt for the 500-unit offer, which includes both business ideas and strategies. There's a 299-unit offer for business ideas only. The speaker plans to eliminate the intermediate offer and introduce a monthly recurring revenue (MRR) option. Philosophically, the speaker is against monthly subscriptions due to potential client "annoyances" (e.g., clients who try the product briefly and then cancel, increasing support load). However, it will be tested with a "somewhat absurd" pricing strategy, possibly 99 per month for monthly and 499 for annual, to avoid chasing "5-euro bills." Product evolution will focus on making existing content more understandable and accessible. Key elements like problem/solution, economic model, go-to-market, and financial projections are good but need clearer presentation. Market trend data is crucial, as many businesses fail because they create something nobody cares about. The "money, relationships, health" aspects, market/sub-market/niche, and the "tech novelty vs. value" matrix will remain. The matrix categorizes ideas into: "unique tech novelty, low value" (attractive but hard to monetize), "boss of the game" (highly unique, high value, captures significant value), "essential basic" (not unique but high value, often representing good opportunities with less competition), and "low impact" (low uniqueness, low value – to be avoided). The focus is on "boss of the game" and "essential basic" opportunities. Competitive analysis and understanding the customer avatar are fundamental. Many businesses fail because they don't know who they're selling to or what problem they're solving. The pain point, current customer behaviors to solve the problem, and the customer's level of awareness (problem-aware, solution-aware, product-aware) are critical. The speaker emphasizes targeting customers who are at least problem-aware or solution-aware, as convincing someone they have a problem when they don't is too difficult. Frustrations with competing solutions and prejudices about the niche are also important considerations. The psychological matrix (fears, dreams, desires, frustrations) helps in understanding customer motivations. Identifying "first adopters" is key for starting a business, often by monopolizing a small niche before scaling (e.g., Facebook starting at Harvard). With AI, geographical boundaries become less relevant, allowing for global targeting of first adopters, though payment issues in certain countries must be considered. The value proposition and minimum viable product (MVP) are emphasized. The speaker criticizes the common "cowardice" of aspiring entrepreneurs who delay launching a product until it's "perfect." He recalls launching IDES V1 with shame because it was "dog shit" but it allowed for feedback and iteration, leading to V2. The V3 will be developed similarly, prioritizing launch over perfection. Future functionalities include optimizing customer lifetime value (LTV) through additional offerings and extended customer engagement. Copywriting, big idea market propositions, and "Dream 100" distribution (identifying key influencers) are also vital. The current IDES platform's "Dream 100" section needs improvement, including adding photos and breaking down influencers by language. Finally, the V3 will incorporate "prompts" for AI, allowing users to automate tasks more effectively. A V4, potentially skipping V3, would focus on "skills" – specific actions or competencies that AI "players" (like OpenCloud, Code, etc.) can execute. The speaker reiterates that the biggest problem for people today is not knowing "the next big thing" (like dropshipping or selling PDFs in 2016) but rather finding niche, rare ideas that align with their personal constraints (budget, time, ambition). He concludes that finding the right idea through IDES can solve financial problems and lead to a good life.