{"slug": "ian-welsh-argues-western-ai-investors-back-losing-bets", "title": "Ian Welsh Argues Western AI Investors Back Losing Bets", "summary": "Blogger Ian Welsh reported switching from Claude to DeepSeek V4 after finding that his monthly AI costs dropped from $3,000-$5,000 to roughly $5 per week. Welsh cited DeepSeek's pricing of $0.28 input and under $1 output per million tokens, compared to Claude's $5 input and $25 output per million tokens. The cost disparity raises questions about the competitiveness of Western AI investments against cheaper Chinese and open-source alternatives.", "body_md": "Photo: \nianwelsh.net\n \n· rights & takedowns\nAccording to Ian Welsh's blog post on IanWelsh.net, the author compares token costs and performance between US-hosted models and the Chinese service DeepSeek V4. The post reports pricing figures: \nClaude\n at \n$5\n input and \n$25\n output per million tokens, \nDeepSeek\n at \n$0.28\n input and under \n$1\n output per million tokens (with discounts), and a reported cached cost of \n$0.0002\n per million tokens. Ian Welsh reports personal spending of \n$3,000\u0013$5,000\n per month on Claude Code before switching to DeepSeek and says his work now costs about \n$5/week\n versus an asserted \n$1,000/week\n for the same workload on Claude. The post argues that Chinese AI offerings, combined with open-source models and efficiency, create a competitive advantage over Western models.\nWhat happened\nAccording to Ian Welsh's blog post on IanWelsh.net, the author describes switching from \nClaude\n to \nDeepSeek V4\n after observing large cost differences. The post gives pricing comparisons: \nClaude\n at \n$5\n input and \n$25\n output per million tokens, \nDeepSeek\n at \n$0.28\n input and under \n$1\n output per million tokens (with current discounts), and a reported cached cost of \n$0.0002\n per million tokens. The author reports previously spending \n$3,000\u0013$5,000\n per month on Claude Code and now paying about \n$5/week\n on DeepSeek, contrasted with a claimed \n$1,000/week\n for the same workload on Claude.\nEditorial analysis\nIndustry-pattern observations: cost-per-token and caching efficiency materially change economics for high-volume, token-intense workloads. Organizations evaluating cloud vs self-hosting often find that aggressive caching and lower inference costs reduce marginal price pressure and can enable different trade-offs between latency, control, and total cost of ownership.\nIndustry context\nIndustry-pattern observations: public commentary emphasizing low-cost, open-source models reflects a broader debate on whether proprietary frontier models retain value when near-parity performance is available at far lower running cost. Analysts and practitioners have repeatedly noted that cheaper inference shifts the levers for product design, deployment scale, and operational budgets.\nWhat to watch\nFor practitioners: monitor independent benchmarks for quality parity, reproducible latency and cost measurements for comparable workloads, and availability of permissive licenses or self-hosting tooling that materially reduce vendor lock-in. For investors and builders: track whether cost reductions are sustained at scale and whether open-source projects maintain active maintenance and security practices.\nScoring Rationale\nThe post highlights a material cost argument that is relevant to practitioners evaluating self-hosting and vendor lock-in. The piece is an opinionated synthesis with no independent benchmarking beyond the author's claims, so relevance is moderate rather than industry-shaking.\nPractice interview problems based on real data\n1,500+ SQL & Python problems across 15 industry datasets — the exact type of data you work with.\nTry 250 free problems", "url": "https://wpnews.pro/news/ian-welsh-argues-western-ai-investors-back-losing-bets", "canonical_source": "https://letsdatascience.com/news/ian-welsh-argues-western-ai-investors-back-losing-bets-45581445", "published_at": "2026-05-28 05:30:50.466807+00:00", "updated_at": "2026-05-28 05:30:53.615699+00:00", "lang": "en", "topics": ["artificial-intelligence", "large-language-models", "ai-products", "ai-tools", "ai-infrastructure"], "entities": ["Ian Welsh", "Claude", "DeepSeek V4", "Claude Code", "DeepSeek", "IanWelsh.net"], "alternates": {"html": "https://wpnews.pro/news/ian-welsh-argues-western-ai-investors-back-losing-bets", "markdown": "https://wpnews.pro/news/ian-welsh-argues-western-ai-investors-back-losing-bets.md", "text": "https://wpnews.pro/news/ian-welsh-argues-western-ai-investors-back-losing-bets.txt", "jsonld": "https://wpnews.pro/news/ian-welsh-argues-western-ai-investors-back-losing-bets.jsonld"}}