AI Integration Transforms Enterprises, Disrupts Industries, and Redefines Online Content
TL;DR
- Agent orchestrators have become a common enterprise skill and product, with major companies like ServiceNow and Salesforce creating dedicated roles and positions for overseeing specialized AI agents.
- Companies are increasingly treating AI agents as digital full-time employees, reclassifying compute costs as capital expenditures and effectively hiring agents for procurement and scheduling roles.
- Enterprises are prioritizing the collection and curation of company-specific unstructured data and decision-making processes to train and enhance reasoning models, leading to a 320-fold increase in reasoning token consumption.
- Professional services firms face a pricing crisis as AI drives fundamental disruption, shifting from billable hours to output-based pricing and enabling up to 40% cost reductions for early adopters.
- Open-source AI models have temporarily surpassed proprietary models on key benchmarks, with several open-source options consistently ranking among the top performers, indicating a significant shift in AI development.
- The traditional ad-supported internet model is under threat as AI answer engines abstract content, leading to a potential future where most new online content is generated for LLMs and scrapers rather than human consumption.
- Social media platforms exacerbate deepfake problems, with fraud incidents surging and the public's ability to distinguish fake media from real media dropping to near zero, enabling widespread misinformation and impersonation.
Deep Dive
The rapid evolution of AI necessitates a continuous re-evaluation of predictions, with a year in AI time equating to roughly seven dog years. This review of January 2025 AI predictions reveals a high degree of accuracy, underscoring the accelerating pace of AI's impact on industries and job roles. The implications extend beyond technological advancements to fundamental shifts in business models, workforce composition, and the very nature of online information consumption.
AI agent orchestrators have emerged as a recognized and growing position within enterprises, with companies like Service Now and major tech firms like Walmart, KPMG, and Salesforce creating roles such as orchestration engineer and AI agent architect. This signifies a structural shift in IT and operations, where specialized agents are managed by human oversight, indicating a growing need for skills in AI workflow management. Similarly, the concept of AI agents being "hired" is gaining traction, with companies like AWS and Klarna reclassifying agent compute costs as digital full-time employees, treating them as labor assets for reporting. Salesforce's Agent Force further blurs the lines by allowing companies to hire agents via subscription, mirroring traditional freelance models. This trend implies a future where human and AI workforces are increasingly integrated and managed under similar operational frameworks.
The importance of company-specific reasoning data for training advanced AI models has been validated, with OpenAI reporting a 320-fold increase in reasoning token consumption. Red Hat's use of synthetic data generation to convert proprietary information into training examples for private reasoning engines highlights a strategic imperative for businesses to curate their unique intellectual property for AI. This suggests that organizations that effectively capture, clean, and categorize their proprietary decision-making data will gain a significant competitive advantage as reasoning models become more prevalent. Furthermore, the professional services sector is undergoing a significant pricing crisis driven by AI, shifting from billable hours to output-based pricing. McKinsey reports up to 40% cost reductions for early AI adopters, leading to substantial job losses in fields like consulting, accounting, and law, as AI can now perform tasks previously done by junior professionals, such as data analysis, trend identification, spreadsheet creation, and PowerPoint generation.
The discourse around Universal Basic Income (UBI) has entered mainstream conversations, partly fueled by concerns over AI-driven job displacement, with proposals and pilot programs emerging in the U.S. This indicates a societal grappling with the potential economic fallout of widespread automation and the need for new social safety nets. Open-source AI models have experienced a surge, temporarily surpassing proprietary models on various benchmarks, demonstrating the rapid democratization of advanced AI capabilities. While proprietary models often reclaim leadership, the continued strong performance of open-source alternatives suggests a more competitive and accessible AI development landscape. Chinese AI, particularly in the open-source domain, is a significant global force, accounting for 30% of usage and causing market confusion through cost reporting discrepancies. Companies like Perplexity have demonstrated strategic agility by pivoting from an "answers engine" model to agentic browsing and specialized applications like shopping and finance, acknowledging the difficulty of competing solely on information retrieval against dominant AI platforms.
API prices for advanced AI models have plummeted, dropping below 50 cents per million tokens, making cutting-edge AI technology more affordable and accessible for a wider range of applications. Embodied AI, encompassing robotics, autonomous vehicles, and drones, is experiencing explosive growth with significant venture capital funding, indicating a future where AI is increasingly integrated into the physical world, impacting logistics, transportation, and last-mile delivery. While AI video generation has not yet fully achieved the one-shot, five-minute HD video capability predicted, the ability to effectively stitch together shorter clips and the significant investment by entities like Disney in AI video suggest that AI-generated content will become increasingly prevalent and potentially surpass human-created media in the coming years. The traditional ad-supported internet model is facing an existential threat as AI answer engines abstract content and an increasing percentage of online content is AI-generated, potentially making the web an archaic interface for accessing information. Finally, social media platforms are exacerbating deepfake problems, making misinformation and impersonation significantly harder to detect, with deepfake fraud incidents surging and a very low percentage of individuals able to distinguish real from fake media, posing a growing threat to individuals and organizations alike.
Action Items
- Audit AI agent job postings: Analyze 5-10 recent postings for AI agent roles to identify common requirements and responsibilities.
- Track AI reasoning token consumption: Measure monthly reasoning token usage for 3-5 core company workflows to identify optimization opportunities.
- Evaluate professional services pricing models: Compare current billable hour rates against output-based pricing for 2-3 service offerings.
- Measure AI-generated content prevalence: Analyze 10-15 newly published web pages to estimate the percentage of AI-generated content.
- Assess deepfake detection accuracy: Test 5-10 internal users on identifying AI-generated media to gauge current detection capabilities.
Key Quotes
"a lot can change in a year think about it in your own personal or professional life how much has it changed in 2025 probably a lot but probably not as much as ai because when we're talking about ai time is like warp speed i mean a year in ai time is like seven dog years so that's like 49 real life years so when i look back at the 2025 ai predictions show that i made back in january of this year it feels like a legit lifetime ago because that's how fast ai moves and how much the landscape shifts almost on a daily basis"
The speaker, Jordan Wilson, highlights the accelerated pace of change within the field of artificial intelligence. He uses the analogy of "warp speed" and "seven dog years" to emphasize how quickly AI evolves, making a year in AI feel like decades in human terms. This rapid evolution necessitates frequent reassessment of predictions and strategies.
"number 25 in january 2025 was agent orchestrators will be a growing position and just fyi i'm probably i'm going to try not to say this every single prediction but think back think back to january 2025 this is when the main ai models were gpt4o and claude 35 sonnet and gemini 2 not gemini 3 not gemini 2 5 gemini 2 right we i mean we had entire ai product lines that hadn't even yet been announced and have already been killed off right open ai's operator hadn't even been announced yet and it's been announced and essentially now killed off and replaced right so keep that in mind when i'm talking about these predictions because when i talk about agent orchestrators right this is the first prediction here my prediction was that it's going to be an actual growing position and that you're going to have people at companies who are going to be essentially agent orchestrators overseeing dozens of specialized agents sounds safe now right at the time i literally couldn't find anything on the internet about agent orchestrators"
Wilson recounts his 2025 prediction about the rise of "agent orchestrators" as a growing job position. He emphasizes that at the time of the prediction, this concept was virtually unknown and unsearchable online, despite the existence of major AI models like GPT-4o and Claude 3.5 Sonnet. This underscores how quickly emerging AI roles can become mainstream.
"number 22 from our january 2025 prediction series high end professional services will go through a pricing crisis so i said ai will drive down prices and reshape billing in law consulting and accounting i'd say definitely got this one right but it is going to be i think an even bigger wave happening here over the next quarter or two because if i'm being honest this one really didn't start to quote unquote hit home until the last two months right some of these ai predictions and in roadmaps hit instantly this one didn't start to hit until literally probably quarter four if i'm being honest"
Wilson discusses his prediction that high-end professional services, such as law, consulting, and accounting, would face a pricing crisis due to AI. He states that this prediction was accurate but notes that its impact was delayed, only becoming significantly apparent in the last two months prior to the recording, contrasting with other predictions that had an immediate effect.
"number 20 open source models surge and they temporarily overtake proprietary models yeah i literally said it will temporarily overtake proprietary models and guess what happened well exactly that happened so you know one of the biggest ones was the deepseek shock so that was in january deepseek released r1 which was an open weights model and at the time the best model in the world was gpt 40 and on many benchmarks the open source model beat open ai's gpt 40 obviously open ai i believe it was just a couple weeks later released an update to gpt 40 that on most benchmarks you know they kind of reclaimed that top spot but this literally happened temporarily open models surge and i think they continue to surge right if you look at the top 10 models i believe right now on lm arena you have depending on the day anywhere from three to four of them are open source"
The speaker reviews his prediction that open-source AI models would surge and temporarily surpass proprietary models. He cites the example of DeepSeek's R1 model outperforming OpenAI's GPT-4o on several benchmarks in January, illustrating how open-source advancements can briefly lead the field before proprietary models catch up. Wilson notes that open-source models continue to hold strong positions in rankings.
"number 17 i said api prices are going to be like early 2000s rap and they were going to drop it like they're hot so my prediction in january which people scoffed at that i said you're going to have frontier or near frontier model api prices drop hard and get below 50 cents per million tokens and everyone's like jordan you're an idiot you have no clue what you're talking about that's you know more than 15x cheaper than what it's currently at not possible oh look yeah i was right sorry"
Wilson discusses his prediction that API prices for advanced AI models would drop significantly, becoming less than 50 cents per million tokens. He notes that this prediction was met with skepticism, with many considering it impossible given the then-current pricing. He asserts that this prediction proved correct, with prices falling dramatically below expectations.
"number 14 the future of traditional internet comes into question so my prediction there was the traditional internet model which is ad supported click through traffic would come into serious question as ai answer engines abstract content away from publishers not only that but just the traditional internet is just going to become slap right and i think we've seen that so a lot of these studies i'm talking about here i don't believe them but i do think that they signify the truth right so there's you know these studies that say you know 90 of new content online will be ai generated by 2026 yeah i'd probably believe that but you can't technically prove it there was an a hrefs study in april that found that 74 of the 900 000 newly created webpage webpages contained ai generated content"
Wilson reviews his prediction that the traditional, ad-supported internet model would be challenged by AI answer engines. He explains that these AI systems abstract content, potentially reducing traffic to publishers. Wilson references studies indicating a high percentage of AI-generated content on new webpages, suggesting a significant shift in how information is accessed and consumed online.
Resources
External Resources
Books
- "State of Enterprise AI 2025" (OpenAI) - Mentioned as a report revealing a 300 and 20-fold increase in reasoning token consumption per organization over 12 months.
Articles & Papers
- "AI is reshaping how McKinsey makes money" (Business Insider) - Referenced as an article indicating fundamental pricing model disruption for professional service firms.
- "Service Now" (MIT Sloan) - Mentioned in relation to having positions in agent orchestration.
- "Deepfake fraud incidents" (iProve) - Discussed as a study finding that only 0.1% of participants could correctly identify all fake and real media.
- "Chinese AI open source models" (OpenRouter study) - Referenced for data indicating 30% of global AI usage is from Chinese open source models.
- "Ahrefs study" (April) - Mentioned for finding that 74% of newly created webpages contained AI-generated content.
- "NPR declared this an extinction level event for online publishers" (July) - Referenced as a declaration regarding the impact of AI on online publishers.
- "Thompson Reuters report" (April) - Discussed as a report stating that professional services pricing is undergoing a recognition of new technologies impacting billing models.
- "McKinsey professional services report" - Referenced for stating that professional service firms are experiencing fundamental pricing model disruption.
People
- Elon Musk - Mentioned as a tech leader who has publicly endorsed UBI.
- Sam Altman - Mentioned as a tech leader who has publicly endorsed UBI.
- Marco Rubio - Mentioned as a US Secretary of State whose voice was cloned by an AI imposter.
Organizations & Institutions
- OpenAI - Mentioned in relation to its "State of Enterprise AI 2025" report and its Sora technology.
- ServiceNow - Mentioned as having positions in agent orchestration.
- Walmart - Mentioned as debuting new AI job titles.
- KPMG - Mentioned as debuting new AI job titles.
- Salesforce - Mentioned as debuting new AI job titles and for its Agent Force platform.
- Adobe - Mentioned as having jobs for Adobe Agent Orchestrator and for its investment in OpenAI.
- AWS - Mentioned for reclassifying agent compute costs as digital FTEs.
- Klarna - Mentioned for reclassifying agent compute costs as digital FTEs.
- Red Hat - Mentioned for detailing how enterprises use synthetic data generation for private reasoning engines.
- Apple - Mentioned as having considered acquiring Perplexity.
- Google - Mentioned in relation to its Gemini models and ViO technology.
- Anthropic - Mentioned in relation to its Claude models and its partnership with Accenture.
- Microsoft - Mentioned as a company that has partnered with Everyday AI.
- Nvidia - Mentioned as a company that has partnered with Everyday AI.
- Accenture - Mentioned for signing deals with OpenAI and Anthropic.
- Disney - Mentioned for announcing a $1 billion investment in OpenAI.
- DeepSeek - Mentioned for releasing the R1 open weights model and its V32 model.
- Perplexity - Mentioned as an answers engine that has pivoted to the agentic browser space.
- Apple Intelligence - Mentioned as a feature that Perplexity's iOS app brought AI actions to the phone before.
- Figure AI - Mentioned as a company in the embodied AI sector that exceeded $1 billion in Series C funding.
- Arc AGI - Mentioned in relation to a test where costs were 390 times cheaper.
- US House - Mentioned for a proposal for UBI.
- Cook County - Mentioned for a permanent guaranteed income program.
- New York - Mentioned for a pilot program.
- US Secretary of State - Mentioned as a position that was mimicked by an AI imposter.
- Fortune 100 companies - Mentioned as companies whose CEOs the podcast host speaks with.
- Big Tech companies - Mentioned as companies where people build AI models.
- Small business CEOs - Mentioned as individuals the podcast host speaks with.
- Big Four consulting companies - Mentioned as companies experiencing job losses and pricing model disruption.
Tools & Software
- GPT-4o - Mentioned as a main AI model in January 2025.
- Claude 3.5 Sonnet - Mentioned as a main AI model in January 2025.
- Gemini 2 - Mentioned as a main AI model in January 2025.
- Gemini 2.5 - Mentioned as a main AI model in January 2025.
- Gemini 3 - Mentioned as a main AI model in January 2025.
- Operator (OpenAI) - Mentioned as a product line that had not been announced and was already killed off.
- Agent Force (Salesforce) - Mentioned as a platform allowing companies to hire agents via subscription.
- Comet Browser - Mentioned as Perplexity's agentic browser.
- Kling AI - Mentioned as a top AI video model with an "extend" feature.
- ChatGPT - Mentioned as a tool that companies partner with Everyday AI for training.
- Copilot - Mentioned as an AI operating system and a tool that companies partner with Everyday AI for training.
Other Resources
- AI Agent Orchestrators - Mentioned as a growing position and a common enterprise product and skill.
- AI Agents - Mentioned in relation to companies posting jobs for them.
- Digital FTEs (Full-Time Employees) - Mentioned as a reclassification of agent compute costs by AWS and Klarna.
- Company Reasoning Data - Mentioned as a crucial area for AI builders and strategy leaders.
- Synthetic Data Generation - Mentioned as a method used by Red Hat to convert proprietary data into training examples.
- Universal Basic Income (UBI) - Mentioned as a topic that has become a mainstream kitchen table topic due to AI job displacement.
- Open Weights Model - Mentioned in relation to DeepSeek R1.
- Answers Engine - Mentioned as a business model for Perplexity.
- Agentic Browser Space - Mentioned as a pivot for Perplexity.
- API Prices - Mentioned as dropping significantly.
- Embodied AI - Mentioned as an exploding sector with large-scale funding.
- Humanoid Robots - Mentioned as a significant AI opportunity.
- Autonomous Vehicles - Mentioned as part of the embodied AI sector.
- Drones - Mentioned as part of the embodied AI sector.
- Wearable AI - Mentioned as part of the embodied AI sector.
- AI Video Tools - Mentioned as a sector that is close to producing five-minute HD videos in one shot.
- Advanced Personalized Media - Mentioned as a concept related to AI video generation.
- AI Video - Mentioned as being used by major studios and expected to be prevalent on Disney+.
- Traditional Internet Model - Mentioned as being called into question due to AI answer engines.
- AI Operating System - Mentioned as a future driver of internet interaction.
- Deepfake Problems - Mentioned as being exacerbated by social media platforms.
- Misinformation - Mentioned as a problem worsened by deepfakes.
- Impersonation - Mentioned as a problem worsened by deepfakes.
- CEO Fraud - Mentioned as a type of fraud targeted by deepfakes.
- AI Imposter - Mentioned in relation to mimicking the US Secretary of State.