“I have no special talent. I am only passionately curious.”-Albert Einstein
Associate Professor
Information, Technology & Operations, Gabelli School of Business
Contact: dmolitor@fordham.edu
Show of hands: left, right, or neither?
Yes — recommender systems
Content credit: Data & Society
Definition: AI is an umbrella term for a set of loosely related technologies.
Making decisions about people based on predictions about future behavior or outcomes.
Recommending content, enforcing policies.
The integration of AI with mechanical systems to perform physical tasks.
Generating new content (audio, code, images, text, video) based on description.
Live demo
Example: Predictive AI
Patterns in historical credit-card fraud are used to train an algorithm that scores new transactions in real time — click one to score it.
← Select a transaction to run it through the model
FRAUD SCORE—
Ethical concerns are amplified by inaccuracy.
In most cases, manual scoring rules are just as accurate, far more transparent, and worth considering.
*Dressel & Farid. The accuracy, fairness, and limits of predicting recidivism. Science Advances 2018. (score can be interpreted as slightly better than random)
Note: since 2023, around 2,800 data centers are planned for construction in the US.
US data-center construction spending (not including the compute!) is now overtaking general office construction
Source: US Census Bureau, construction spending (SAAR); US Bureau of Labor Statistics, CPI-U (SA). *Excludes computer equipment; data-center series starts 2014 · hover the chart for monthly values
Class vote
One company's market cap, 2016–today (hover for values). Vote — then we reveal what the dashed line marks.▲ dashed line = ChatGPT launch (Nov 30, 2022)
…but demand still outstrips supply — Nvidia can’t get TSMC to increase capacity fast enough
Source: Intel and Nvidia SEC filings (10-K, 10-Q, XBRL). Plotted at each fiscal quarter’s end date; Nvidia’s fiscal year ends in late January. Some Q4 values are annual revenue less Q1–Q3. Hover for quarterly values
the fuel
the engine
the recipe
Based on a large corpus of text (input), LLMs predict the next word (output).→ a sophisticated version of autocomplete
This toy network: 52 connections (65 parameters with biases).Real LLMs: billions to trillions of parameters — that is why they are called “large” language models.
Framing adapted from Sebastian Raschka’s blog post on building reasoning models: magazine.sebastianraschka.com/p/building-reasoning-models
Easy and obvious use cases of Generative AI
Agents
Coding agents can build tools and analyze files. Personal agents can coordinate tasks across connected apps.
Makes execution easier — allowing more ideas to be pursued.
Source: State of AI in Business 2025 Report (mlq.ai)
Anthropic quarterly revenue ($bn), from its S-1 filing
Source: Anthropic Form S-1 / S-1/A (unaudited quarterly schedule; Q3 2026 from Amendment No. 1, Nov 2, 2026). FY2025 revenue $4.6bn; first nine months of 2026 $33.2bn (+1,206% YoY). Hover a bar for details
Share of U.S. businesses paying for AI tools (%), July 2023 – September 2026
Adoption has more than doubled since mid-2023 — and Anthropic passed OpenAI in May 2026
Source: Ramp AI Index (https://ramp.com/data/ai-index). Hover for monthly values
Source: Epoch AI, Epoch Capabilities Index (CC BY 4.0), retrieved Nov 1, 2026 · 254 models · hover a point for details; click an organization to highlight it
Source: METR, Time Horizon 1.1 (retrieved Nov 1, 2026). Human-expert duration of tasks an AI agent completes at the selected success rate; mostly software, ML and cyber tasks. METR: values above 16 h are unreliable.
Impact of AI: some tasks will be augmented, some will be replaced, and some will be untouched. Moreover, there will be entirely new AI-based tasks.
→ More diverse tasks provide more resilience.* Narrow roles, not so much…
*Especially those requiring judgement, relationships, and context
Example: Consultant
→ Automating some tasks makes the remaining ones more valuable.
https://aleximas.substack.com/p/how-will-ai-driven-automation-actually
*Telehealth provider for GLP-1 drugs
Writing code → managing coding agents (Claude Code, Codex, etc.)
The bottleneck moves to verifying output — coding skills are still required.
https://www.cnbc.com/2026/04/14/ceos-are-betting-ai-will-augment-work-rather-than-displace-all-workers.html
Combines technical & managerial skills.
Flexible capabilities that apply across industries.
Acquire skills along the knowledge frontier in tech & analytics.
Not locked into one career path.
Become a Data Scientist, IT Consultant, or Cybersecurity Specialist.
IT vs. IS: information technology is mainly hardware and software; information systems also include data, people, and processes. It’s critical for managers to think about systems, rather than just technologies, when deploying technology-enabled solutions.
According to BCG: 10% of a company’s efforts should focus on algorithms, 20% on technology and data, and the remaining 70% — the lion’s share — on people and processes.
dmolitor@fordham.edu
GROUNDFLOORGabelli School of Business
Six required courses & four electives (3 tracks)
Four required IS courses
Three required IS courses
…new AI courses for all GSB students (3rd year): AI Analytics [Spring] · AI for Business [Fall]
Python, Systems Analysis, Data Management, Web Apps, Project Management, Capstone
Machine Learning for Business, Advanced Python, Web Analytics, Sports Analytics, etc.
AI and Data-Driven Marketplaces, Mobile Commerce & Apps, Fintech Introduction, etc.
Accounting Information Systems, Cybersecurity, AIS Consulting, etc.
Familiarity with concepts, frameworks, software tools, techniques, and trends in business analytics.
Manage and analyze data to drive management insight, informed decision-making and superior business performance out of raw data.
Drive business transformation through digital business technologies.
In-depth understanding of cutting-edge digital business trends: AI, digital marketplaces, mobile commerce, cloud computing, social technologies, and tech-startups.
Architect, build, and leverage business systems in organizations, including enterprise systems.
Design and manage sophisticated systems that provide the foundations for redesigned business processes, enterprise-wide integration, novel services, and innovative business models.
B.S. in Business Administration with a Concentration in Information Systems
B.S. with a Secondary Concentration or Minor in Information Systems
B.S. in Accounting Information Systems
B.S. in Business Administration with a secondary concentration in Fintech
Technology industry roles ranked among the top jobs for 2025
Source: NACE Salary Survey data (national); *Fordham/Gabelli IS major salaries are actually higher (~)
“Ground truth” data available → models can be optimized
$700bn planned in 2026 for the big four — for comparison, global telecoms is ~$300bn and oil & gas is ~$1tr
Source: company filings and guidance; includes capital/finance leases. *Amazon does not break out AWS capex; AWS is estimated as the vast majority of Amazon’s. Hover a bar for details
Stages of AI
Metz, “OpenAI Scale Ranks Progress Towards Human-Level Problem Solving,” Bloomberg, July 11, 2024. Source: Bloomberg reporting.