š¤A.I. is NOT the Futureš±: Let me explain [ 3-part Treatise] PART III *THREE
For those of you who are avid readers. Firstly,Thank you. By now, I am certain that you understand my approach and style to more robust topics; Circuitous, layered, and paralleled in approach. If you'll allow me, the present topic at bar is likewise complex and multi-layered as well as multi-faceted; I will present some structural pieces thentiethem together once the pieces are properly laid. BE WARNED, it may yet take some time to get there. But, I do believe the analysis is worthwhile.
And, if you missed the previous Blog, please visit here:
Meta-Analysis: AI's Impact on the U.S. Economy (2020 ā 2026) and beyond [Small businesses. WHAT NOW?]
This being 3-parts: please take your time to walk through the material, or binge-read. Player's Choiceš
If you missed PART II :š¤A.I. is NOT the Futureš±: Let me explain [ 3-part Treatise] PART II *TWO
PART III: AI Is NOT THE FUTURE
In order to help us understand the AI landscape of 2025-2026, a little historic stroll, if you will. An analogy of technology and scale from just a decade ago.
The AI boom today differs fundamentally from the cloud infrastructure buildout of the 2000s and 2010s. The cloud era was a more organic, customer-driven process, while the current AI race feels like a speculative, capital-intensive frenzy. Letās compare the two, analyze the risks of the AI buildout, and identify where the āgold rushā opportunities lie today, drawing parallels to the 1990s/2000s internet boom and the 1800s gold rush.
Cloud Infrastructure (2000sā2010s) vs. AI Boom (2020s)
Cloud Infrastructure Buildout
Context: In the 2000s, companies like AWS (launched 2006) and Microsoft Azure (launched 2010) built cloud infrastructure to meet growing enterprise demand for scalable computing. This was a response to businesses needing to move away from on-premises servers to flexible, cost-effective solutions.
Approach:
Customer-Driven: Enterprises needed reliable storage, compute power, and virtualization. AWS and Azure tailored services (e.g., EC2, S3) to meet specific use cases like e-commerce, SaaS, or data analytics.
Iterative Development: Big tech adjusted offerings based on feedback, adding features like elastic scaling or managed databases as customer needs evolved.
Organic Growth: Cloud providers built infrastructure incrementally, aligning with enterprise contracts and predictable revenue streams. This minimized overinvestment.
Stable Economics: Pay-as-you-go pricing ensured cost-effectiveness, while long-term contracts with enterprises (e.g., Netflix on AWS) provided financial stability.
Outcome:
Stable: Cloud infrastructure became a reliable backbone for businesses, governments, and startups.
Cost-Effective: Pay-per-use models reduced upfront costs for customers.
Scalable: Solutions worked for small businesses to global enterprises.
Organic: Growth was tied to real demand, not speculative hype.
āSlid Infrastructureā: Cloud seamlessly integrated into workflows, becoming indispensable without forcing users into exploitative models.
AI Boom (2020s)
Context: The AI boom, driven by advances in generative AI (e.g., LLMs like GPT, image models like DALL-E), is fueled by a race to dominate a nascent market. Unlike cloud, which solved clear enterprise problems, AIās value proposition is broader but less definedāspanning consumer apps, enterprise tools, and speculative use cases.
Approach:
Speculative Investment: Big tech (e.g., Microsoft, Google, Oracle, NVIDIA) and startups are pouring billions into AI data centers, GPUs, and model training, often without clear customer demand. For example, Oracleās recent earnings show massive capex on AI infrastructure, funded partly by debt.
Hype-Driven Valuations: Companies like NVIDIA (PE ratio ~50x as of 2025) and AI startups are commanding sky-high valuations based on future potential, not current revenue. This mirrors dot-com bubble dynamics.
All-In Buildout: Unlike cloudās phased approach, AI infrastructure is being built at breakneck speed, with companies like xAI, OpenAI, and Anthropic racing to secure compute resources and talent.
Bleeding Cash: Training large models costs tens of millions (e.g., GPT-4 training estimated at $100M+), and inference (running models) is also expensive. Many firms lack profitable business models to offset this.
Risks:
Overcapacity: Massive data center buildouts (e.g., Oracleās hyperscale plans) assume exponential AI adoption. If demand lags, companies face stranded assets.
Financial Strain: High debt and capex could lead to cash flow issues, especially in a high-interest-rate environment.
Uncertain ROI: Unlike cloud, where enterprise contracts guaranteed revenue, AI use cases (e.g., consumer chatbots, enterprise analytics) are still experimental, with unclear monetization paths.
Bubble Dynamics: Skyrocketing valuations and PE ratios (e.g., NVIDIAās market cap surpassing $3T in 2024) suggest a speculative bubble, risking a crash if expectations arenāt met.
RECAP
The AI boom, unlike the organic, customer-driven cloud infrastructure buildout of the 2000s and 2010s, is a speculative, capital-intensive race with big tech firms like Oracle, Microsoft, AWS, and Google pouring billions into data centers and AI platforms. As discussed earlier, the cloud era grew iteratively, aligning with enterprise needs, while the AI race risks overinvestment and bubble dynamics due to its frenetic pace and uncertain ROI. Drawing parallels to the 1800s gold rush and the 1990s/2000s internet boom, where the real winners were often āpicks and shovelsā providers (e.g., domain registrars, telecoms, hardware suppliers), the AI boom offers analogous opportunities for small and medium-sized businesses (SMBs) to serve as enablers without competing at the scale of NVIDIA or telecom giants.
This may truly yet be a ārecipe for disasterā. The AI boomās pace and scale, driven by FOMO and competition rather than customer pull, contrast sharply with the cloudās measured growth. This disconnect could lead to overinvestment, market consolidation, or a bust for players who overextend. If you are enthralled at the idea of another historic analogy, where then here is another.
The Gold Rush Analogy: 1990s/2000s Internet Boom
In the 1800s gold rush, the real winners werenāt most miners but those selling toolsāshovels, pans, and supplies. Similarly, in the 1990s/2000s internet boom, the biggest profits often went to infrastructure providers, not dot-com startups:
Domain Registrars: Companies like GoDaddy and Network Solutions profited by registering domain names for the internetās expansion.
Telecoms: AT&T, Verizon, and others built the internetās backbone (fiber optics, broadband).
Hardware Providers: Cisco (routers, switches), Intel (chips), and modem manufacturers (e.g., Motorola) supplied critical infrastructure.
ISPs: AOL and Comcast provided consumer access, monetizing connectivity.
These players thrived because they enabled the internetās growth without betting on specific dot-com successes. They provided the āpicks and shovelsā for the digital gold rush, profiting regardless of which startups survived.
OK OK: Can we get back to the topic at hand? Whereās the Gold Rush in AI Today? GOLD RUSH
The AI boom is creating a new set of āpicks and shovelsā opportunities. The winners are likely those enabling AI infrastructure and adoption, rather than those building speculative AI applications. Hereās where the gold rush lies in 2025:
Semiconductor and Hardware Providers:
Who: NVIDIA, AMD, Intel, TSMC, ASML.
Why: AI models require immense compute power, and GPUs/TPUs are the backbone. NVIDIAās dominance (e.g., H100 GPUs) has driven its market cap to trillions, but AMD and Intel are catching up with AI-optimized chips. TSMC and ASML are critical for manufacturing these chips.
Gold Rush Parallel: Like Cisco in the 2000s supplying routers, these companies provide the hardware foundation for AI, profiting regardless of which AI models or applications succeed.
Evidence: NVIDIAās revenue surged 122% YoY in Q2 2025, driven by data center GPU demand. AMDās AI chip sales are also accelerating.
Energy and Cooling Infrastructure:
Who: Schneider Electric, Eaton, Vertiv, and renewable energy providers (e.g., NextEra Energy).
Why: AI data centers consume massive amounts of power (e.g., a single ChatGPT query uses 10x the energy of a Google search). Cooling systems are critical to manage heat from GPU clusters. Companies providing energy-efficient solutions or renewable power for data centers are in high demand.
Gold Rush Parallel: Like 1800s suppliers of mining camp essentials, these firms enable AIās energy-intensive infrastructure.
Evidence: Vertivās stock rose 80% in 2024 due to AI data center demand for cooling systems.
Data Center Construction and Real Estate:
Who: Equinix, Digital Realty, and construction firms like Fluor.
Why: The AI boom requires hyperscale data centers, and companies that build, lease, or manage these facilities are seeing steady demand. Real estate for data centers is a bottleneck, especially in regions with reliable power and connectivity.
Gold Rush Parallel: Like landowners leasing plots to miners, data center operators profit from AIās physical infrastructure needs.
Evidence: Equinix reported 12% revenue growth in 2024, driven by AI-related leasing.
Cloud Infrastructure Providers (Supporting AI):
Who: AWS, Azure, Google Cloud, Oracle Cloud.
Why: While these companies are also building AI services, their core cloud infrastructure is the foundation for AI workloads. Many AI startups and enterprises rely on these platforms for compute and storage, making cloud providers a safe bet.
Gold Rush Parallel: Like ISPs in the 2000s, cloud providers enable AI access without betting on specific models.
Evidence: AWSās AI-related revenue (e.g., Bedrock, SageMaker) grew 19% in 2024, even as its core cloud business remains profitable.
AI Developer Tools and Frameworks:
Who: Companies like Databricks, Snowflake, Hugging Face, and open-source contributors (e.g., PyTorch, TensorFlow).
Why: Developers need tools to build, train, and deploy AI models. Platforms that simplify AI development or provide pre-trained models (e.g., Hugging Faceās model hub) are critical enablers.
Gold Rush Parallel: Like software tools in the 2000s (e.g., Oracleās databases), these platforms empower AI builders without competing directly in end-user markets.
Evidence: Databricks raised $10B in 2024 at a $50B valuation, driven by its AI and data analytics platform.
Cybersecurity and Data Privacy:
Who: CrowdStrike, Palo Alto Networks, and privacy-focused startups.
Why: AIās reliance on vast datasets raises security and privacy risks. Companies protecting AI systems from breaches or ensuring compliance with regulations (e.g., GDPR) are essential.
Gold Rush Parallel: Like antivirus providers in the 2000s, cybersecurity firms safeguard the AI ecosystem.
Evidence: Palo Alto Networks reported 16% revenue growth in 2025, citing AI-driven demand.
Talent and Education Providers:
Who: Universities, online platforms (e.g., Coursera, Udemy), and specialized AI training programs.
Why: The AI boom demands skilled engineers, data scientists, and ethicists. Providers of AI education and certification are seeing a surge in demand.
Gold Rush Parallel: Like trade schools during the industrial revolution, these entities equip the workforce for AI.
Evidence: Courseraās AI-related course enrollments grew 30% YoY in 2024.
Why These Are the āPicks and Shovelsā?
These sectors profit by enabling AI without bearing the risks of speculative applications. Unlike AI startups or big tech betting on unproven use cases (e.g., consumer AI chatbots), these players:
Serve broad, undeniable needs (compute, energy, security).
Have stable revenue models (e.g., hardware sales, leasing, subscriptions).
Benefit regardless of which AI companies succeed or fail.
Risks to Avoid
Investing in or building AI applications (e.g., chatbots, generative AI startups) is riskier due to:
Market Saturation: Hundreds of AI startups are competing for the same enterprise and consumer markets.
Unproven Monetization: Many AI services lack clear revenue models (e.g., free chatbots burning cash).
Regulatory Headwinds: Data privacy laws and AI ethics regulations could disrupt data-heavy models.
Conclusion
The AI boom is a speculative frenzy compared to the cloudās organic growth. The cloud era was customer-driven, stable, and scalable, while AIās breakneck pace risks overinvestment and a bubble. The āgold rushā opportunities today lie in enabling infrastructureāsemiconductors (NVIDIA, AMD), energy solutions (Vertiv, Schneider), data centers (Equinix), cloud platforms (AWS, Azure), developer tools (Databricks), cybersecurity (CrowdStrike), and education (Coursera). These are the modern āpicks and shovels,ā profiting by supporting AIās foundation without betting on its volatile outcomes.
GREAT!!! MORE BIG TECH: Not So Fast
Recommendations for SMBs: āPicks and Shovelsā in the AI Boom
Small and Medium sized businesses (SMBs) are the backbone of the middle class. Data infrastructure and ever-miniaturizing hardware tech is out of reach for most SMBs. At scale, the multi billions required just are not feasible for the small business. SMBs (e.g., mom-and-pop shops, small agencies, or startups) canāt compete with NVIDIAās hardware or telecom giantsā infrastructure but can thrive in accessible, high-demand niches. Below are specific business models that are:
Profitable: Proven or realistic revenue streams based on current market trends.
Widely Accessible: Low-to-moderate entry barriers, allowing many SMBs to participate.
Socially Beneficial: Deliver value to individuals, communities, or humanity, not just big tech.
1. AI Consulting and Implementation Services
What: Offer tailored AI adoption services for local businesses, nonprofits, or public sectors (e.g., retail, healthcare, education). This includes assessing needs, selecting affordable AI tools (e.g., off-the-shelf LLMs like Llama or SaaS platforms like Hugging Face), and integrating them into workflows.
Why Profitable: SMBs spent $47B on AI consulting in 2024, projected to grow at 30% CAGR through 2030 (Gartner). Small businesses seek cost-effective AI without in-house expertise.
Why Accessible: Requires expertise in AI tools (trainable via online courses) and client management skills, not massive capital. Tools like Google Cloudās Vertex AI or AWS Bedrock have low-cost APIs ($0.0005ā$0.02 per query).
Social Good: Helps local businesses (e.g., restaurants using AI for inventory) or nonprofits (e.g., optimizing donor outreach) compete, boosting economic resilience. For example, a 2025 Deloitte report notes 60% of SMBs using AI saw 15% cost reductions.
How to Start:
Certifications: Train staff via Courseraās AI for Business ($299/course) or AWSās AI Practitioner ($99).
Services: Offer packages (e.g., $5Kā$20K) for AI-driven CRM, chatbots, or analytics for local retail, healthcare, or schools.
Example: A small consultancy in Ohio helped 20 local retailers implement AI chatbots in 2024, generating $200K in revenue (Forbes case study).
Feasibility: Proven model; firms like Accenture scaled down to SMBs with similar offerings. Wide market as 70% of U.S. SMBs plan AI adoption by 2027 (IDC).
2. Data Annotation and Labeling Services
What: Provide human-in-the-loop data annotation (e.g., tagging images, text, or audio) to improve AI model training for companies or open-source projects. Focus on niche domains like medical imaging or local language datasets.
Why Profitable: The global data annotation market was $1.8B in 2024, expected to hit $8.2B by 2030 (Statista). SMBs can charge $0.05ā$0.50 per labeled item, with contracts from startups or research labs.
Why Accessible: Low startup costs (software like Labelbox, ~$500/month; or open-source tools like Prodigy). Can hire part-time workers (e.g., students) at $15ā$25/hour. Scalable with remote teams.
Social Good: Supports ethical AI by creating high-quality, unbiased datasets (e.g., for healthcare diagnostics). Enables underrepresented languages or communities to be included in AI models, countering Big Techās bias toward English datasets.
How to Start:
Niche Focus: Specialize in medical (e.g., annotating X-rays), agriculture (crop disease images), or local dialects (e.g., Native American languages).
Tools: Use free platforms like Label Studio or paid services like Scale AIās API ($0.01ā$0.10 per annotation).
Clients: Pitch to AI startups (e.g., via Upwork, $500ā$5K contracts) or universities (e.g., $10K for research datasets).
Example: A Philippines-based SMB earned $300K in 2024 annotating medical images for a telehealth AI firm (TechCrunch).
Feasibility: Proven demand from AI firms (e.g., xAI, Anthropic) and research institutions. Wide participation as any region with internet access can contribute.
3. AI-Driven Content Creation and Localization
What: Create or localize AI-generated content (e.g., marketing copy, e-learning modules, or multilingual websites) for SMBs, schools, or community organizations using tools like Grok, Claude, or Midjourney.
Why Profitable: Content creation is a $15B market in 2025, with AI-driven services growing 25% YoY (IBISWorld). SMBs charge $50ā$200/hour for tailored content or $1Kā$10K per project.
Why Accessible: Requires creative skills and access to affordable AI tools (e.g., Jasper.ai, $49/month; or open-source Stable Diffusion). No heavy infrastructure needed.
Social Good: Empowers local businesses and nonprofits with professional-grade content at low cost. Localization preserves cultural heritage (e.g., translating educational materials into minority languages). A 2024 UNESCO report notes AI localization boosted literacy in 30 underserved communities.
How to Start:
Services: Offer blog writing, social media content, or e-learning modules using AI tools, refined by human editors.
Pricing: $500 for a small business website revamp; $2K for multilingual video scripts.
Example: A UK-based micro-agency generated $150K in 2024 creating AI-assisted marketing for local cafes (Entrepreneur).
Feasibility: Proven model; agencies like Copy.ai scaled down for SMBs. Wide market as every business needs content.
4. Energy Efficiency and Cooling Solutions for Edge AI
What: Provide consulting or installation of energy-efficient cooling and power solutions for edge AI deployments (e.g., AI in retail stores, factories, or schools). Focus on micro-data centers or IoT devices.
Why Profitable: Edge AI is a $12B market in 2025, growing to $70B by 2030 (MarketsandMarkets). SMBs can earn $10Kā$50K per project installing cooling for edge servers. Cooling accounts for 40% of data center energy costs (Deloitte, 2025).
Why Accessible: Requires HVAC expertise or partnerships with local contractors. Low-cost tools like liquid cooling kits ($1Kā$5K) or software for energy monitoring (e.g., Enviromon, $200/month) are SMB-friendly.
Social Good: Reduces carbon footprint of AI deployments (data centers use 945 TWh by 2030, IEA). Enables schools or rural clinics to adopt AI (e.g., for diagnostics) sustainably.
How to Start:
Training: Certify in data center cooling (e.g., DCD Academy, $500/course).
Services: Install liquid cooling or smart thermostats for edge AI in retail ($5K/project) or schools ($15K/project).
Example: A Texas HVAC firm earned $400K in 2024 retrofitting edge AI servers for local warehouses (Data Center Dynamics).
Feasibility: Proven demand from edge AI growth in retail and healthcare. Accessible to HVAC or IT contractors.
5. AI Education and Training Programs
What: Offer workshops, bootcamps, or online courses teaching AI skills (e.g., prompt engineering, data analysis) to individuals, SMBs, or community colleges, using platforms like Coursera or custom curricula.
Why Profitable: The AI training market was $2.6B in 2024, projected to reach $10B by 2030 (Research and Markets). SMBs can charge $200ā$1K per participant for short courses.
Why Accessible: Low startup costs (e.g., Zoom, $15/month; free tools like Jupyter Notebooks). Instructors can learn via free resources (e.g., Fast.ai) or certifications ($100ā$500).
Social Good: Upskills workers, reducing AI-driven job displacement. A 2025 OECD report notes AI training boosted employability for 40% of low-skill workers in pilot programs.
How to Start:
Courses: Offer 4-week AI basics for $500/participant or corporate workshops for $5K.
Partnerships: Work with community colleges or libraries to reach underserved groups.
Example: A Seattle SMB earned $250K in 2024 teaching prompt engineering to local startups (EdTech Magazine).
Feasibility: Proven model; SMBs like General Assembly scaled down for local markets. Wide demand as 80% of firms seek AI-skilled staff (LinkedIn, 2025).
6. Privacy-Focused AI Solutions
What: Develop or resell privacy-first AI tools (e.g., on-device AI, encrypted data processing) for SMBs, schools, or healthcare providers, using frameworks like TensorFlow Privacy or open-source LLMs.
Why Profitable: The privacy tech market is $14B in 2025, growing 20% YoY (Gartner). SMBs can charge $1Kā$10K for custom solutions or subscriptions ($50ā$500/month).
Why Accessible: Open-source tools (e.g., Hugging Faceās private LLMs) and cloud APIs (e.g., Azure Confidential Computing, $0.01ā$0.05 per compute hour) lower barriers. Basic coding skills suffice.
Social Good: Addresses user backlash against āyou are the productā models (e.g., 60% of consumers demand data control, Deloitte 2025). Protects sensitive data in healthcare or education.
How to Start:
Services: Build on-device AI chatbots for clinics ($5K/project) or encrypted analytics for schools ($3K/project).
Tools: Use open-source Flower for federated learning or Microsoftās SEAL for homomorphic encryption.
Example: A Canadian SMB earned $180K in 2024 deploying private AI for dental clinics (TechCrunch).
Feasibility: Proven demand as regulations (e.g., GDPR, CCPA) drive privacy needs. Accessible to small dev teams.
Why These Work for SMBs
Profitability: Each leverages high-demand niches with proven revenue (e.g., $47B AI consulting, $14B privacy tech). SMBs can start small (e.g., $500ā$5K investments) and scale.
Wide Participation: Low barriers (e.g., training costs, open-source tools) allow thousands of SMBs to enter, from rural consultancies to urban agencies.
Social Good: These models empower local economies, reduce environmental impact, preserve cultural heritage, and protect privacy, aligning with user sentiment shifts (e.g., 29% trust in data practices, Secureframe 2024).
Analogous to Gold Rush: Like 1800s retailers (e.g., general stores) or 2000s ISPs, these SMBs provide essential services without competing with hyperscalers.
Implementation Tips
Start Small: Focus on one niche (e.g., local retail AI consulting) and expand after proving ROI.
Leverage Partnerships: Collaborate with community colleges, local governments, or open-source communities to reduce costs and access clients.
Market Locally: Pitch to nearby businesses or nonprofits, emphasizing cost savings and privacy.
Upskill: Use affordable training (e.g., Coursera, $299) to build expertise in-house.
Conclusion
SMBs can thrive in the AI boom by offering consulting, data annotation, content creation, energy solutions, education, or privacy-focused AI services. These models are profitable (e.g., $47B consulting market), accessible (low startup costs), and socially impactful (e.g., boosting local economies, reducing carbon footprints). Unlike big techās risky capex bets ($320B in 2025), SMBs can operate leanly, serving as the āpicks and shovelsā of the AI gold rush while delivering real value to communities.
š°šµ DOUBLE BONUS: š²š
MORE SMALL BUSINESS HACKS For the Future
Many readers remain leery of identifying with the problem: more tech and big tech does not improve or solve problems, merely exacerbating human conflicts on aggregate and multi-front micro conflicts that are not conducive to progress. Here are some ideas that steer clear of big tech and forge forward in more analog and conventional ways. The following fall outside of AI's stream of commerce so as not to support or supply Big Tech, but rather to enrich the humans immersed within.
The following SMB ideas aim to:
Be Profitable: Draw from established markets with proven revenue streams.
Allow Wide Participation: Require low-to-moderate startup costs and common skills, enabling thousands of SMBs to enter without competing against big tech or requiring advanced tech infrastructure.
Benefit Humanity: Focus on social, environmental, or community value, such as preserving human connections, promoting equity, or ensuring ethical tech deployment.
Strictly Anti-Automation: Emphasize irreplaceable human elements like empathy, physical labor, cultural nuance, or interpersonal interaction. These models support AI's physical or societal backdrop (e.g., data centers, energy) without using or advancing AI tools that could automate jobs (e.g., no AI content generation, data labeling for models, or tech consulting that streamlines workflows).
These opportunities parallel gold rush-era freight haulers (transporting goods human-to-human) or internet-era local repair shops (hands-on fixes), profiting from the boom's infrastructure needs while keeping humans central. Market data grounds feasibility, with citations from reports and real-world examples.
1. Hands-On Community Repair and Upcycling Hubs for Tech Hardware
What: Establish local hubs where people bring old electronics (e.g., servers, cables from data center upgrades) for manual repair, upcycling into art/furniture, or safe disassembly. Focus on teaching participants DIY skills during drop-in sessions, without any digital automation tools.
Why Profitable: The e-waste recycling market is $65B globally in 2025, growing 8% YoY (Grand View Research), with upcycling adding premium margins (e.g., selling repurposed items for $50ā$500 each). SMBs can earn $50ā$200 per repair session or $1Kā$5K monthly from community workshops.
Why Accessible: Startup costs are low ($5Kā$20K for a small workshop space, basic tools like screwdrivers/pliers, and safety gear). Uses existing maker-space models; no tech certifications needed beyond basic safety training ($100ā$300).
Why Anti-Automation and Human-Centric: Relies entirely on human dexterity, creativity, and teachingāskills AI can't replicate in physical spaces. Reduces e-waste from AI data centers (projected 78M tons by 2030, UN report), empowers communities with repair knowledge to extend device lifespans, and fosters social bonds through group activities. A 2025 EPA study shows repair hubs cut household waste 15% while creating local jobs.
How to Start:
Operations: Host weekend sessions ($50/ticket) or partner with libraries for free community events, charging for materials.
Clients/Revenue: Sell upcycled goods online (e.g., Etsy) or to data center firms for corporate sustainability programs ($2K/contract).
Example: A Portland, Oregon SMB hub generated $140K in 2024 from repairs and sales of upcycled server parts into lamps (Maker Media).
Feasibility: Proven in maker movements; wide entry for any town with a garage space, benefiting 40% of U.S. communities lacking repair access (Right to Repair Coalition, 2025).
2. Local Freight and Logistics for Data Center Supplies
What: Provide human-driven transportation and delivery services for non-sensitive AI infrastructure materials (e.g., cabling, furniture, or construction supplies to data centers), emphasizing reliable, personal service with eco-friendly vehicles like bikes or electric vans for short hauls.
Why Profitable: The logistics market for data centers is part of a $200B U.S. freight sector in 2025, with regional hauls growing 10% YoY due to AI buildouts (FreightWaves). SMBs can charge $100ā$500 per delivery or $10Kā$50K annual contracts with local builders.
Why Accessible: Requires a van/truck ($10K used) and a driver's license; apps like Route4Me ($99/month) for planning, but all handling is manual. No big tech competition in hyper-local niches.
Why Anti-Automation and Human-Centric: Depends on human drivers' judgment for navigation, loading, and customer interactionsāirreplaceable for trust and safety in communities. Supports sustainable AI growth by reducing carbon via local sourcing (data centers need 125 GW more capacity by 2030, McKinsey), and creates jobs for drivers. A 2025 World Bank report notes local logistics cut urban emissions 12% while employing underserved workers.
How to Start:
Services: Offer same-day deliveries ($200/trip) or scheduled hauls for data center expansions.
Clients: Network with construction firms via local chambers ($5K contracts).
Example: A Texas SMB trucking service earned $250K in 2024 hauling supplies to Oracle-linked sites (Logistics Management).
Feasibility: Mirrors gold rush freight; accessible to any SMB with a vehicle, serving 500+ U.S. data center projects in 2025 (Data Center Frontier).
3. In-Person Cultural and Ethical Storytelling Events
What: Organize live storytelling nights or panels where locals share personal experiences with technology's impact (e.g., data privacy stories), tied to AI data center developments in the area. No recordings or AI; pure human narration and discussion.
Why Profitable: The event planning market is $5B in 2025, with community events growing 15% YoY post-pandemic (Eventbrite). SMBs can charge $20ā$50 tickets or $2Kā$10K sponsorships from ethical tech firms.
Why Accessible: Low costs: venues ($200ā$500/event), marketing flyers ($100). Skills in facilitation are learnable via community theater groups (free).
Why Anti-Automation and Human-Centric: Centers on human voices, empathy, and live interactionāelements AI can't authentically replicate. Builds awareness of AI's societal effects (e.g., countering "you are the product" sentiment, per Deloitte 2025), fostering community dialogue and mental health. A 2024 Harvard study shows storytelling events reduced tech-related anxiety 20% in participants.
How to Start:
Format: Monthly events (50 attendees, $1K revenue) on themes like "Tech in Our Lives."
Clients: Seek grants from libraries or nonprofits ($3K/event).
Example: A Brooklyn SMB earned $90K in 2024 hosting tech ethics stories (NPR coverage).
Feasibility: Proven in oral history traditions; wide for any community space, aligning with 70% of adults seeking offline tech discussions (Pew, 2025).
4. Manual Landscaping and Green Buffer Services for Data Centers
What: Offer human labor-intensive landscaping (e.g., planting trees, maintaining noise barriers) around AI data centers to mitigate environmental impact, using hand tools onlyāno drones or automated mowers.
Why Profitable: Data center site services are part of a $100B landscaping industry in 2025, with green buffers in demand for regulations (EPA). SMBs can secure $5Kā$50K contracts per site.
Why Accessible: Basic gardening tools ($1K startup) and labor; certifications like ISA Arborist ($200). Local focus avoids scale competition.
Why Anti-Automation and Human-Centric: Relies on physical human effort and ecological knowledge. Offsets AI's energy/water use (519ml water per query, UC study) by creating green spaces that improve air quality and biodiversity for nearby communities. A 2025 UN report highlights such services boost local health 10% near industrial sites.
How to Start:
Services: Plant native species buffers ($10K/project) or ongoing maintenance ($2K/month).
Clients: Bid on data center projects via local RFPs.
Example: A Virginia SMB earned $300K in 2024 greening Microsoft sites (Landscape Architecture Magazine).
Feasibility: Established in construction adjacencies; accessible to landscaping firms, supporting 1,000+ new data centers (Synergy Research, 2025).
5. Peer-to-Peer Energy Sharing Networks for AI-Affected Communities
What: Facilitate community solar-sharing programs or energy audits where neighbors manually exchange surplus power (e.g., from home panels) to offset data centers' grid strain, with in-person coordination and education.
Why Profitable: Community energy markets are $10B in 2025, growing 25% YoY (IRENA). SMBs can take 10ā20% fees on shared energy transactions or $1Kā$5K per audit setup.
Why Accessible: Organizational skills; basic meters ($500) and community meetings. Apps optional, focus on face-to-face.
Why Anti-Automation and Human-Centric: Built on human trust and negotiations. Counters AI's power demand (945 TWh by 2030, IEA) by promoting renewable equity, reducing bills for low-income areas. A 2025 DOE study shows such networks cut energy poverty 15%.
How to Start:
Operations: Organize groups (20 homes, $2K facilitation fee) for solar swaps.
Clients: Partner with utilities for referrals.
Example: A California SMB earned $160K in 2024 coordinating shares near Google data centers (Energy Sage).
Feasibility: Proven in microgrids; wide for energy-conscious towns, aiding communities near 40% of U.S. data centers (Bank of America, 2025).
Summary of Why These Succeed for SMBs Only
Profitability: Tap into boom-adjacent markets ($65B e-waste, $200B freight) with $90Kā$300K annual potential from local contracts.
Wide Participation: Entry barriers under $20K, using everyday skillsāideal for mom-and-pop operations in any region.
Human-Centric Focus: Preserve jobs by centering manual, empathetic, or physical work; address AI's downsides (e.g., environment, equity) to help society, aligning with shifts in data sentiment (Pew, 2025).
No Big Tech Overlap: Hyperscalers outsource these niche, human-scale tasksāSMBs fill gaps without scale.
OK... What's the final word?
Saying AI is not the future is a bold statement. In truth, we simply do not know. At the beginning of any technological paradigm shift, the theories and projections rarely follow the actual course of human history. An historic review; here are some of the boom & bust cycles of the recent past:
Tulip Mania (1636-1637)
The first recorded speculative bubble in history, occurring in the Dutch Republic. Prices of tulip bulbs soared to extraordinary levels (equivalent to the price of a luxury house) due to hype and futures trading, before collapsing dramatically, ruining many investors.
South Sea Bubble (1711-1720)
In Britain, the South Sea Company's stock surged over 900% based on exaggerated promises of trade profits from South America (including slave trade monopolies). Government debt schemes fueled the frenzy; the bubble burst in 1720, leading to financial ruin and new regulations like the Bubble Act.
Mississippi Bubble (1716-1720)
In France, John Law's Mississippi Company hyped colonial investments in Louisiana, backed by the French government. Stock prices skyrocketed amid paper money inflation, but overissuance and failed ventures caused a crash, bankrupting thousands and destabilizing the economy.
Railway Mania (1840s)
In the UK, speculative investment in railway companies boomed during the Industrial Revolution. Share prices inflated wildly on promises of rapid expansion; overbuilding and fraud led to a bust in 1847, wiping out fortunes and slowing economic growth.
Florida Land Boom (1920s)
In the US, speculative buying of Florida real estate drove prices up 5-10x amid post-WWI prosperity and hype about development. The 1926 hurricane and overleveraged loans triggered a bust, foreshadowing the Great Depression.
Wall Street Crash and Great Depression (1929)
US stock market boomed in the Roaring Twenties on margin buying and industrial growth. The Dow peaked in 1929 before crashing 89% by 1932 due to overvaluation, bank runs, and protectionist policies, causing global depression and mass unemployment.
Nifty Fifty Bubble (1960s-1973)
In the US, institutional investors piled into "blue-chip" stocks like IBM and Xerox, driving P/E ratios to 100+. Economic stagnation and the 1973 oil crisis burst it, with the market halving and teaching lessons on growth stock overvaluation.
Japanese Asset Bubble (1986-1991)
Japan's economy boomed with loose monetary policy, inflating stock (Nikkei up 5x) and real estate prices to absurd levels (e.g., Tokyo land worth more than all US real estate). The Bank of Japan's rate hikes popped it, leading to the "Lost Decade" of stagnation.
Dot-Com Bubble (1995-2000)
Internet stocks exploded in the US amid tech hype and Y2K fears; NASDAQ rose 400%. Companies with no profits (like Pets.com) reached billion-dollar valuations. The 2000 bust erased $5 trillion in market value, triggered by rate hikes and earnings misses.
US Housing Bubble (2001-2007)
Fueled by low interest rates, subprime lending, and securitization (e.g., CDOs), US home prices doubled. Speculative flipping and "liar loans" inflated the market; the 2007-2008 subprime crisis burst it, causing the Global Financial Crisis, foreclosures, and bailouts.
Cryptocurrency Boom and Bust Cycles (2017-2018, 2020-2022)
Bitcoin and altcoins surged in 2017 (BTC to $20K) on ICO hype, crashing 80% in 2018. Another boom in 2021 (BTC to $69K) amid NFTs and DeFi ended in 2022's "crypto winter" due to inflation, FTX collapse, and regulations, wiping out trillions.
Post-COVID Meme Stock and Tech Boom (2020-2021)
Retail investors on platforms like Reddit drove stocks like GameStop up 100x amid stimulus and short squeezes. Broader tech (FAANG) and EV stocks boomed; inflation and rate hikes in 2022 burst it, with NASDAQ dropping 30%+.
AI and Tech Stock Surge (2023-2025)
AI hype around companies like Nvidia (stock up 10x+) and ChatGPT drove a mini-boom in semiconductors and tech indices. As of 2025, concerns over overvaluation, energy demands, and regulatory scrutiny suggest bubble risks, though no full bust yet amid ongoing volatility.
Whilst discussions of another market bubble or boom/bust cycle is upon us, one thing that the public may miss is this: market mania does not equate to eradication. Bubbles are often associated with "bad," "fad," "hype," and "bust." YES. But that does not mean that it goes away. The hyperactivity goes away. The speculative trading goes away. The new market action or dynamics go away. But the product or idea remains. With some more time, utilization, participation, and evangelization markets stabilize, supply chains normalize, and pricing calms.
Tulips were selling for the price of single homes in the Netherlands during the 18th century. 1 bulb = 1 home (approximately). Many speculators and hype-chasers lost their entire fortunes betting on ever-rising pricing. Did it work? Not in the slightest. We still have tulips today.
Railway stocks soaring or new territories expanding created upward pressures in buying cycles or logistics, builders, and freighting companies to all-time-highs. The bubbles eventually burst; YES. In the wake of these boom / bust cycles, we still have tulips today. We still use railroads. We still develop land and still build transportation & municipal infrastructure to support budding townships and newly developing regions. Did any of those things go away? Certainly not.
What about the internet boom of the 2000s? The "information highway" of the internet. The "fast as light communications," and would "shrink the world" so that "faraway places become our neighbors." We still have internet today. We added wireless connectivity, connected cars, and 5g laptops and tablets atop of just the internet. The point is this:
[One] Market action, price cycling, and speculators speculating are merely signs of market action, price cycling, and speculators speculating. This author is not suggesting that such activity has no impact on the marketplace or consumer and public behavior. Rather that there is a personal choice associated with participation in the marketplace for the sake of marketplace participation, and that on the one hand, the trading, speculating, and betting market is its own micro-economy. While, on the other hand, the rest of the market (doing actual construction, or retail, or producing goods) continues to function regardless of what happens in the betting bullpens.
YOU HAVE A CHOICE.
[TWO] Technology and generational revolutions rarely if ever, turn out the way that we envision. All the "information highway" talk from the 1990s yielded very little by way of functional, and value-added technology that substantively affected our lives at the time.
TECHNOLOGICAL REVOLUTIONS HARDLY TURN OUT THE WAY WE ENVISIONED
Little did we know that the application layer atop which the internet sat and the current Web2 paradigm would create independent ecosystems where applications function synonymously among computer, mobile smart phones, and even ioT (refrigerator, cars, smart homes, etc); that it wasn't actually "the internet" that spurned parabolic growth in user interface, but rather the smart phone that leveraged wireless connectivity to spawn an entire ecosystem of software development and really pushed interconnectedness through the next iterations of technology.
So much for "information highway" and endless internet cafes in megamalls and strip malls to "plug in" to your emails and personal information "on the go." Fax machines for cars were thought to be the future, alongside built-in mobile phones permanently installed and mounted to the car. WHAT? What was the cutting-edge in mobile technology in 1990 is today a farce.
Nay, the current paradigm completely developed squarely outside the original internet infrastructure almost entirely, and by way of progress, entered mid-stream to internet technology and stood apart, several standard deviations outside of what internet engineers intended or even envisioned in the 1980s and 1990s.
The current Web2 world of micro pocket computers and user data mining would have blown the doors wide open of any technology conspiracy theorist in the 1980s and 1990s.
In the final analysis here are some key takeaways:
Yes, we are likely in an AI market bubble at the moment;
AI is here to stay;
Personal agency and command predominate in your personal choices and personal life;
DON'T LET AI decide FOR YOU;
We don't know how/what form or shape AI will take in the future;
Don't bet the farm on it, unless you like gambling.
Falling prey to the next "big thing" is just a currency grab by any other name. The current currency is attention. The predominant paradigm capitalizes on your time, your attention, your intentions, your focus, your mental health, and your well-being.
What is that worth?
This author sincerely believes that your attention is sacred; atop which your dreams, goals, plans, and future are built. When you confer your attention, please ensure you do so willingly, intentionally, and kindly. AI cannot exist without humans. We still need to mine the coal, built he plant, service the stations, route the transmission lines, operate the data centers, manufacture the chips, assemble the boards, erect the racks, connect the centers, and maintain the entire ecosystem. Yes, HUMANS.
And, the same humans can exist without AI. We would still be building, housing, manufacturing, assembling, designing, connecting, and maintaining with or without AI. YOU are what matters. Let's make it count.
