In the bustling halls of the Bloomberg New Economy Forum in Singapore, Matthew Fitzpatrick, CEO of Invisible Technologies, sat down with Bloomberg Television hosts Shery Ahn and Avril Hong to unpack the future of AI. Amid swirling debates about an AI bubble, Fitzpatrick offered a grounded perspective on technology’s trajectory, his company’s recent funding triumph, and the gritty realities of enterprise adoption. This interview, aired on Bloomberg, comes at a pivotal moment as AI investments soar, yet real-world implementation lags.

Invisible Technologies, a key player in AI infrastructure, recently raised $100 million in growth funding, valuing the company at over $2 billion, according to reports from Bloomberg and BusinessWire. Led by Vanara Capital, a TPG spinoff, this round underscores investor confidence in Invisible’s hybrid model that blends human oversight with AI training. Fitzpatrick, who joined as CEO in early 2025 after leading McKinsey’s AI efforts, as detailed in a BusinessWire announcement, emphasized that the funds are fueling core technology platforms focused on enterprise needs.

Navigating the Bubble Debate

Fitzpatrick addressed bubble concerns head-on, distinguishing between technological advancements and valuation froth. “If you ask the question about a bubble you can think about it in two ways. The technology itself and valuations. And they are separate questions,” he told Bloomberg Television. He highlighted a “step change in model performance” over the past four years, asserting the paradigm shift is “very real.” Yet, on valuations, he noted that Big Tech’s earnings make their high multiples defensible, while AI funding concentrates in a handful of private firms—about 100 companies capturing 70% of investments.

This scarcity of capital around elite AI players contrasts with past bubbles, as Fitzpatrick pointed out that public software multiples are half of 2021 levels, per his interview. Recent sentiment on X echoes this nuance; one post from investor Bindu Reddy predicted an AI bubble burst in mid-2026 due to overhyped data center investments totaling $7 trillion, while another from moninvestor argued against a bubble, citing cash-flow funding over credit hype. These views align with Reuters reporting on Big Tech earnings under the “specter of AI bubble.”

Funding Surge and Strategic Investments

The $100 million raise, initially announced but later upsized due to demand, as Fitzpatrick revealed, is being channeled into platforms for AI training and validation. “We are investing it in our core technology platforms,” he explained on Bloomberg. Invisible has trained over 80% of the world’s top AI models, serving giants like Microsoft and Amazon Web Services, according to the same outlet. This positions the company as a rival to Scale AI, with a focus on enterprise-grade precision.

Fitzpatrick described the enterprise cycle mirroring model builders’ processes: creating rubrics, scoring outputs, and ensuring reliability. For instance, generating credit review documents requires rigorous testing—”You need to do all the same stuff that we’ve been doing for the model builders the last couple of years,” he said. This investment comes amid broader trends; Lucidworks‘ 2025 AI Benchmark Study, based on insights from 1,600+ AI leaders, projects agentic AI as a 2026 priority, highlighting adoption gaps in clean data and production scaling.

Challenges in Enterprise Adoption

Adoption at the enterprise level remains “difficult,” Fitzpatrick admitted. A KPMG report cited in the interview shows 60% of people worldwide use generative AI monthly, indicating robust consumer uptake. However, an MIT study reveals only 5% of AI projects reach production. “You need to have clean data. You need to have testing and validation,” he stressed, warning that enterprises seek “extreme precision” and human-like quality in outputs.

Imagine betting your annual bonus on a thousand AI-generated memos being accurate—that’s the stakes, per Fitzpatrick. This resonates with X posts; one from Giorgi Orbeliani noted McKinsey data showing two-thirds of companies stuck in AI experimentation, despite $611 billion in projected 2026 infrastructure spend. Similarly, Yahoo Finance coverage of Invisible’s funding emphasizes its role in fixing enterprise AI through human-in-the-loop models.

The Data Dilemma and Market Scale

Data quality is paramount: “Rubbish in and rubbish out,” as host Avril Hong put it. Fitzpatrick agreed, predicting the data labeling market will quadruple by 2030. “We are in the first inning of that,” he said on BizToc, which summarized the interview. Enterprises are just beginning to tackle unstructured data like contracts, conversations, and videos, while fragmented structured data in CRM systems poses additional hurdles.

“My favorite joke is that when good AI meets bad data, the data usually wins,” Fitzpatrick quipped. This underscores the need for organization before AI deployment. X user KY highlighted 2025’s $400 billion U.S. AI infrastructure investments yielding only $50 billion in revenue, forecasting a 2026 shift from hype to results—mirroring Pulse 2.0‘s take on Invisible’s funding for hybrid solutions.

Sector-Specific Opportunities

Investment has flowed into expected sectors like financial services and healthcare, which boast large tech organizations. But Fitzpatrick sees untapped potential elsewhere. In agriculture, AI can analyze video and image data for crop yields and herd safety. Consumer goods, like SwissGear luggage, benefit from inventory forecasting analytics powered by Invisible’s platforms.

Sports offer exciting applications: tracking spatial movement patterns for player drafting, as Invisible did for the Charlotte Hornets. “Everyone loves sports,” Fitzpatrick noted. This diversification sets Invisible apart from competitors like Scale AI, where training is only a third of its business. “We are a tech-centric platform,” he emphasized, serving eight sectors with an enterprise focus.

Competitive Edge and Future Outlook

Differentiating in a crowded field, Invisible leverages its broad enterprise services beyond pure AI training. As StartupHub.ai reports, the $100 million bet is on hybrid models, not full automation. X posts from investors like Chris Camillo predict big money flowing into AI subsectors like embodied AI by 2026, while Hedgie warns of energy bubbles tied to AI hype.

Fitzpatrick’s vision for 2026 centers on bridging adoption gaps. With Big Tech’s AI capex projected at $600 billion by Wedbush’s Ives, as shared on X by Byul, Invisible aims to enable precise, scalable deployments. As one X post from AR Insider notes, investment exceeds current demand, but Fitzpatrick’s pragmatic approach—rooted in data integrity and testing—positions Invisible to thrive amid evolving trends.

Investor Sentiment and Broader Implications

Amid debt-financed deals emerging, as flagged by X user Ted Zhang citing Citrini7, the AI landscape teeters between innovation and overvaluation. Yet, Fitzpatrick’s stats on concentrated funding and private dynamics suggest resilience. Bloomberg‘s coverage reinforces this, with Invisible claiming a stake in training most global AI models.

Looking ahead, enterprises must conquer data fragmentation and validation hurdles. Fitzpatrick’s insights, drawn from Invisible’s work with top clients, highlight a path forward: invest in platforms that ensure AI outputs are not just generated, but trusted. As X user Gerard predicts a correction wiping out weak AI startups by 2026, companies like Invisible, with their enterprise-centric strategies, may emerge stronger.

Strategic Priorities for 2026

The Lucidworks study forecasts gaps in agentic AI adoption, urging strategic priorities like clean data pipelines. Invisible’s funding, per Yahoo Finance, targets exactly this, powering next-gen infrastructure. Fitzpatrick’s McKinsey background informs this focus, as noted in his appointment announcement.

In agriculture and sports, Invisible’s applications demonstrate AI’s transformative potential beyond finance. With consumer adoption surging, enterprises lag—but Fitzpatrick’s roadmap, emphasizing precision and human vigilance, could accelerate progress. As AI infrastructure spend balloons, Invisible’s model offers a blueprint for sustainable growth.