AI Business with Fexingo: Artificial Intelligence Companies, Models, and Enterprise Adoption

Lucas and Luna examine an overlooked shift in the AI business landscape: the sudden obsession with gross margins. As AI model commoditization accelerates, companies like NVIDIA, Broadcom, and Super Micro Computer are seeing diverging stock performance tied directly to margin narratives. The hosts break down why Broadcom's 12.3% weekly gain reflects its custom-chip margin story, while NVIDIA's flat...

Episode 29 of AI Business with Fexingo examines the growing divergence between AI hardware stocks and big tech shares as of June 2026. Lucas and Luna analyze data showing ARM Holdings up 24% in a week, Broadcom up 14%, while Alphabet and Amazon have dropped 7-9%. They explore three drivers: the shift from AI training to inference chips, hyperscaler capex rotations away from consumer cloud, and the...

Uber capped employee AI spending after blowing through its budget in just four months. Lucas and Luna break down what happened, why it's not a sign of AI disillusionment, and what it reveals about the real cost of enterprise AI adoption. They examine how the shift from pilots to production is hitting budgets hard, why companies like Uber are imposing guardrails instead of cutting AI entirely, and...

Lucas and Luna break down the latest surge in AI hardware stocks, with a spotlight on SMCI, ARM, and AVGO. They discuss the shift from training to inference driving demand for custom chips, the data center buildout, and why enterprise adoption is finally translating into real revenue for semiconductor companies. Plus, a look at how Google's new fake call detection fits into the AI security landsca...

Alphabet just announced plans to raise $80 billion for its AI infrastructure buildout. Lucas and Luna break down what that staggering number actually buys — data centers, GPUs, networking, power — and why it signals a shift in how Big Tech thinks about AI spending. They contrast Alphabet's capex-heavy approach with Microsoft's more modular strategy and Nvidia's pivot to the CPU market with AI agen...

As of June 1, 2026, AI hardware and semiconductor stocks are seeing massive gains: Palantir up 19.2%, Broadcom up 9%, and Super Micro up 28.8% in just five days. In this episode, Lucas and Luna unpack what's driving the rally—from enterprise inference demand to the IPO filing by Anthropic. They focus on the shift from training to inference hardware, the rise of custom ASICs, and why investors are...

Episode 24 of AI Business with Fexingo: Lucas and Luna unpack the accelerating shift from custom-built AI models to third-party APIs, using live market data—Broadcom up 7.9% in five days, AMD surging 10.4%—as a signal. They trace the logic through a concrete case: how a mid-size logistics provider cut its ML budget 40% by swapping an in-house LLM for an API. The hosts explore why enterprises are h...

Lucas and Luna dive into a growing concern in AI circles: model accuracy improvements are slowing even as companies pour more compute into training. They discuss recent benchmarks showing GPT-5's accuracy gains over GPT-4 were smaller than GPT-4's over GPT-3.5, and how NVIDIA's latest GPU sales to hyperscalers may reflect this plateau. The hosts explore what this means for enterprise AI adoption—i...

Lucas and Luna explore how the AI model market is rapidly commoditizing, with major implications for enterprise buyers and investors. Drawing on recent market moves—SoftBank's €75 billion French data center bet, Meta's reported AI pendant, and the backlash against GitHub Copilot's token billing—they argue that differentiation is shifting from model quality to distribution, data moats, and user exp...

GitHub Copilot just switched to token-based billing, and developers are furious. Lucas and Luna break down why the new pricing model sparked so much backlash, what it means for enterprise AI adoption, and how Microsoft is navigating the tension between monetization and developer trust. They discuss the specific numbers behind the outcry, compare Copilot's approach to competitors like Amazon CodeWh...

Episode 20 of AI Business with Fexingo dives into the growing trend of developers refusing to code without AI assistance. Lucas and Luna discuss a recent TechCrunch headline about coders who feel AI tools are indispensable, and how this dependency could backfire on careers. They explore the rise of AI-assisted coding platforms like GitHub Copilot, the risk of skill atrophy, and what it means for e...

Episode 19 of AI Business with Fexingo dives into a quietly tectonic shift in enterprise AI: the retreat from custom model building. Lucas and Luna unpack why companies like JPMorgan and Walmart are abandoning in-house models in favor of third-party APIs, using fresh data from May 2026 — including Super Micro Computer's 30% weekly surge and Broadcom's 5% gain — to illustrate the pivot from trainin...

Episode 18 of AI Business with Fexingo looks at the paradox hitting enterprise AI in mid-2026: companies are slashing AI budgets even as total spending climbs. Lucas and Luna dig into a new data point from Gartner showing that 47% of enterprises now have a formal AI cost-reduction program, up from 12% a year ago. They explore why this is happening — from GPU over-provisioning to failed proof-of-co...

Episode 17 of AI Business with Fexingo. Lucas and Luna explore why Paris has quietly emerged as the most important AI hub outside Silicon Valley. They break down the ecosystem's secret sauce: a deep bench of elite math talent from schools like Polytechnique and CentraleSupélec, combined with a regulatory sandbox that France has actively built. The hosts discuss how Mistral AI raised €600 million w...

A Google engineer is charged with insider trading after allegedly making $1.2 million on Polymarket using confidential AI launch data. Lucas and Luna explore how this case exposes the weak compliance controls inside Big Tech's AI race, why Polymarket's prediction-market model creates new insider-trading risks, and what the SEC's enforcement action means for the AI industry's culture of secrecy. Th...

Lucas and Luna explore the disconnect between massive AI startup funding rounds and the slow pace of real enterprise adoption. They discuss today's news that AI coding startup Cognition raised $1 billion at a $25 billion pre-money valuation, and drill into why enterprises struggle to move beyond pilot projects. The hosts examine the 'last mile' problem in AI deployment, the gap between vendor prom...

On this episode of AI Business with Fexingo, Lucas and Luna explore how the shift to AI-powered search is quietly eroding user privacy. DuckDuckGo's installs are up 30% as users reject being 'force-fed' Google's AI Search, a trend that signals growing discomfort with how AI models scrape and store personal data. The hosts break down the trade-offs: better answers vs. more surveillance, and what en...

A TechCrunch report from earlier today reveals a surprising trend: Indian gig workers are being recruited to train AI models for physical robots. Lucas and Luna dig into why this matters for enterprise AI adoption, the economics of data labeling at scale, and what it means for the future of automation. They tie it to recent moves by companies like Microsoft and Tesla in the AI training space, and...

Enterprise AI adoption is booming, but most teams are stuck on the 'last mile' — integrating model outputs into real workflows. Lucas and Luna dig into a concrete example: a Fortune 500 retailer that spent millions on an LLM-powered inventory system, only to see 40% of its recommendations ignored by warehouse managers. They discuss why fine-tuning and infrastructure investment haven't solved the i...

Lucas and Luna dive into why the biggest AI bottleneck in 2026 isn't model capability—it's the 'last mile' of integration into messy enterprise workflows. Using the ClickUp layoff as a jumping-off point, they explore how companies are struggling to turn raw AI horsepower into actual productivity gains. Specific examples from Palantir's deployment playbook and the surprising role of API latency in...

Lucas and Luna explore a major shift happening right now in enterprise AI: the move from training massive models to running inference at scale. They unpack why AMD is suddenly beating NVIDIA in this new phase, using real data from the past week — AMD up 11 percent while NVIDIA slipped 3 percent. They discuss how inference changes the economics of AI for companies, from lower hardware costs to ener...

This episode of AI Business with Fexingo dives into the massive power consumption driving today's AI boom — and how it's reshaping everything from data center design to utility stocks. Lucas breaks down the numbers behind training a single large language model, which can consume as much electricity as a small town in a year. Luna questions whether the industry can sustain this trajectory without t...

Ferrari's F1 team has partnered with IBM to deploy an AI system that transforms telemetry data, race strategy, and driver history into personalized fan experiences. Lucas and Luna break down how the system works, why it matters for enterprise AI adoption, and what it signals about the future of brand engagement. With Ferrari's stock up 3.7% this week and a 50-basis-point surprise from the Fed, the...

In this episode of AI Business with Fexingo, Lucas and Luna unpack the quiet revolution reshaping enterprise AI: open-source models. While NVIDIA's stock slips 3.1% over the past five days, Meta's Llama 3 and other open-weight models are flooding corporate deployments, forcing a rethink of the AI stack. Lucas cites Gartner data predicting 65% of enterprises will use open-source AI by 2027, and the...

AMD just posted an 11.1% weekly gain while NVIDIA slipped 3.1%. Lucas and Luna dig into the AI chip market split between training and inference, and why AMD's MI400 series is winning real workloads from cloud giants like Microsoft and Meta. They look at AMD's new ROCm 6.2 software stack and how enterprise buyers are diversifying away from NVIDIA. Plus: why the inference market might be bigger than...