Larry Ellison and the Graveyard of "Hot Shit Tech"

# Larry Ellison and the Graveyard of "Hot Shit Tech"

**Author:** Patrick Duggan

---

**Post 18. In the 90's, Oracle databases were the hottest shit in enterprise tech. Now Larry Ellison has 2% cloud market share and got TikTok as a political gift. This is the inevitable fate of all "hot shit tech."**

The 90's: Oracle Was EVERYTHING

**1990-2000: Oracle Database dominated enterprise:**

- 60%+ market share

- $100K+ per CPU licensing

- Every Fortune 500 company ran Oracle

- Larry Ellison became 5th richest person on Earth

- **Status:** If you weren't on Oracle, you weren't serious

**Larry's yacht collection grew with the licensing fees.**

**Oracle OpenWorld conferences:** 60,000+ attendees worshipping at the altar of relational databases.

**The message:** Relational databases are the ONLY way to handle enterprise data. Everything else is a toy.

2025: Larry's Not Even in the Conversation

**Oracle Cloud (2025):**

- 2% market share

- AWS: 32%

- Azure: 23%

- Google Cloud: 11%

**How Larry got TikTok:**

- Political gift from Trump administration (not competitive win)

- "Partnership" to host TikTok data

- Zero technical innovation involved

- Pure regulatory capture

**Oracle's relevance today:**

- Legacy enterprises too scared to migrate

- Database licensing revenue (milking existing customers)

- Cloud strategy: distant 4th place

- **Status:** Not even in the conversation for startups

The Pattern: Hot Shit Tech → Graveyard

**This happens EVERY decade.**

1990's Hot Shit Tech (Now Dead/Irrelevant)

| Tech | Peak Status | 2025 Status |

|------|-------------|-------------|

| **Oracle DB** | 60% market share | 2% cloud, legacy only |

| **Netscape Navigator** | 80% browser dominance | Discontinued 2008 |

| **RealPlayer** | Streaming media pioneer | Nobody remembers |

| **Yahoo Directory** | $125B market cap (2000) | Sold for $5B (2017) |

| **Flash** | 99% browser penetration | Killed 2020 |

| **GeoCities** | 38M users (1999) | Shut down 2009 |

| **AltaVista** | Top search engine | Closed 2013 |

| **AOL** | $160B market cap (2000) | Verizon bought for $4.4B |

**The pattern:**

**Year 0:** "This is the future! Everyone must use this!"

**Year 5:** Market dominance, licensing fees, conferences

**Year 10:** Competitors emerge with better tech

**Year 15:** "Legacy system" status

**Year 20:** Historical footnote

**Year 25:** "Grandpa, what's Oracle?"

Vince Gilligan's Meta-Commentary

**Breaking Bad taught us this lesson through Walter White's lab equipment.**

Season 1: High School Chemistry Lab

- Outdated equipment

- Embarrassing, ancient tech

- Works, but looks pathetic

- Walter's shame: "I'm a high school teacher with beakers from 1987"

Season 5: State-of-the-Art Methylamine Lab

- Cutting-edge equipment

- Gleaming stainless steel

- Looks like a pharma facility

- Walter's pride: "This is REAL chemistry"

The Punchline:

**Both labs make the same 99.1% pure meth.**

**The equipment changed. The chemistry didn't.**

**Hot shit tech gets replaced. Quality work remains constant.**

Oracle = Walter's High School Beakers

**1990's Oracle:**

- State-of-the-art relational database

- Cutting-edge SQL engine

- Enterprise-grade reliability

- **Walter's Season 5 lab energy**

**2025 Oracle:**

- Legacy relational database

- Ancient SQL licensing model

- Vendor lock-in nightmare

- **Walter's Season 1 beakers energy**

**The irony:**

Oracle databases STILL WORK. They're reliable. Battle-tested. Proven.

But they LOOK old. FEEL old. Cost too much. Lock you in.

**Like Walter's high school beakers: functional but embarrassing.**

AWS Replaced Oracle (But AWS Will Be Replaced Too)

**2006: AWS launches**

- "Cloud databases? That's not serious enterprise tech."

- Oracle laughs

- Larry says: "Cloud is just someone else's computer"

**2015: AWS dominates**

- 32% cloud market share

- RDS, Aurora, DynamoDB eating Oracle's lunch

- Startups born on AWS, never touch Oracle

**2025: AWS is the new Oracle**

- Vendor lock-in (egress fees, proprietary APIs)

- Expensive (cost optimization = full-time job)

- Metadata leakage (Walmart avoids AWS for this reason)

- **Status:** Still dominant, but showing cracks

**2035: AWS will be the "old hot shit tech"**

**Guaranteed.**

The Graveyard Tour

Let's visit the corpses of "essential" tech that's now irrelevant:

Stop 1: Netscape Navigator ($4.2B IPO → $0)

**1995:** Netscape IPO births dot-com boom

- 80% browser market share

- Marc Andreessen on magazine covers

- "The internet IS Netscape Navigator"

**1998:** Microsoft bundles Internet Explorer with Windows

- Antitrust lawsuits

- Market share collapses

- AOL buys Netscape for $4.2B

**2008:** Netscape discontinued

- Firefox rises from the ashes

- Chrome kills Firefox

- Netscape = historical footnote

**Lesson:** Bundling beats quality. Microsoft won by making IE free.

Stop 2: RealPlayer (Streaming Pioneer → Forgotten)

**1995:** RealNetworks invents streaming media

- RealAudio, RealVideo

- Required plugin for watching videos online

- 85% market share

**2000:** Windows Media Player + QuickTime compete

- Fragmentation hell (3 required plugins)

- Users hate it

**2010:** YouTube + HTML5 kill all plugins

- No RealPlayer needed

- Streaming = native browser feature

- RealNetworks = "wait, they still exist?"

**Lesson:** Platforms beat plugins. Native beats third-party.

Stop 3: Yahoo ($125B → $5B Fire Sale)

**1994-2000:** Yahoo Directory = how you find websites

- $125B market cap (January 2000)

- Rejection of Google's acquisition offer ($1M in 1998)

- "Portal strategy" (news, mail, search, everything)

**2000-2008:** Google eats their lunch

- Search becomes dominant

- Yahoo's directory model obsolete

- Multiple failed CEO transitions

**2017:** Verizon buys Yahoo for $4.4B

- 96% value destruction

- Merged with AOL (two corpses in one coffin)

- Email service for people who can't figure out Gmail

**Lesson:** Aggregation beats curation. Algorithms beat human editors.

Stop 4: Flash ($3.4B Acquisition → Security Nightmare)

**1996:** Macromedia Flash revolutionizes web animation

- Interactive websites

- Online games

- Video players (YouTube used Flash until 2015)

**2005:** Adobe buys Macromedia for $3.4B

- 99% browser penetration

- Required for "rich internet applications"

- Web designers worship Flash

**2007:** Steve Jobs refuses to support Flash on iPhone

- "Flash is buggy, insecure, battery drain"

- Adobe fights back

- iPhone wins

**2020:** Flash officially discontinued

- HTML5 replaces everything Flash did

- Security holes finally closed

- Archive.org preserves Flash games as historical artifacts

**Lesson:** Proprietary formats die. Open standards win (eventually).

Why Hot Shit Tech Always Dies

**Three reasons:**

1. Vendor Lock-In Breeds Resentment

**Oracle's licensing model:**

- Per-CPU pricing ($47,500 per core)

- Annual maintenance fees (22% of license cost)

- Audit threats ("we think you're using more cores")

- Exit costs (rewrite entire application)

**Result:** Customers HATE you. They use your product because they're trapped, not because they want to.

**When an alternative appears (AWS RDS), they flee.**

2. Market Dominance Breeds Complacency

**Yahoo in 2000:**

- "We're the portal! We have everything!"

- Search, news, mail, games, chat, GeoCities

- "Nobody can compete with our scale"

**Google in 2000:**

- "We just do search. And we're really good at it."

**Result:** Focused competitor beats bloated incumbent.

3. Technology Moves Faster Than Businesses

**Netscape's mistake:**

- Sold browsers for $49

- Microsoft gave IE away for free

- Netscape couldn't pivot fast enough

**Flash's mistake:**

- Proprietary plugin model

- Mobile shifted to native apps

- HTML5 standardized what Flash did

- Adobe couldn't pivot fast enough

**Oracle's mistake (current):**

- Relational databases for everything

- Cloud shifted to NoSQL, object storage, serverless

- Oracle launched cloud too late

- Can't compete with AWS/Azure on price

**Result:** Fast-moving startups kill slow-moving incumbents.**

The Vince Gilligan Lesson

**Breaking Bad's central theme:**

**Technology changes. Quality doesn't.**

**Walter White:**

- Season 1 beakers → Season 5 methylamine lab

- Equipment changed

- 99.1% purity didn't change

- **Quality was constant**

**Jesse Pinkman:**

- Chili powder meth (Season 1) → Blue meth (Season 5)

- Recipe changed

- Craftsmanship improved

- **But Walter's obsession with purity was always there**

**The lesson for tech:**

Oracle's technology (relational databases) will be replaced.

But the PRINCIPLES Oracle proved (ACID transactions, data integrity, reliability) remain valuable.

**Hot shit tech dies. Foundational principles survive.**

What Replaces Oracle?

**Current "hot shit" (will be obsolete in 10 years):**

Serverless Databases

- DynamoDB, Cosmos DB, Firestore

- "NoSQL is the future!"

- **Prediction:** Niche use cases, not universal replacement

Vector Databases

- Pinecone, Weaviate, Qdrant

- "AI embeddings need vector search!"

- **Prediction:** Absorbed into existing databases (Postgres pgvector)

Graph Databases

- Neo4j, Neptune, TigerGraph

- "Relationships are the future!"

- **Prediction:** Niche use cases (fraud detection, social networks), not universal

**What ACTUALLY replaces Oracle:**

**Boring, reliable, cheap databases that just work.**

Probably Postgres. Because:

- Open source (no vendor lock-in)

- Battle-tested (since 1986)

- Extensions (PostGIS, pgvector, TimescaleDB)

- Community-driven (not profit-driven)

**Postgres is the Commodore 64 of databases:**

- Old

- Boring

- Proven

- Still relevant

**Oracle is the RealPlayer of databases:**

- Once essential

- Now legacy

- Expensive

- Replaceable

DugganUSA's Position

**We use ZERO relational databases.**

Not because we're anti-database. Because we're anti-lock-in.

**Our stack:**

- Flat files (JSON in Azure Blob Storage)

- Redis (7-day TTL, ephemeral data)

- No schema migrations

- No vendor lock-in

- $77/month operational cost

**Is this "hot shit tech"? No.**

**Is this the future? Probably not.**

**Does it work? Yes.**

**Can we migrate to Postgres/DynamoDB/whatever in 35 minutes if needed? Yes.**

**That's the point.**

**We're not betting on the NEXT hot shit tech. We're betting on portability and simplicity.**

The Inevitable Cycle

**Every decade, new "hot shit tech" emerges:**

**1990's:** Oracle, Netscape, RealPlayer, Flash

**2000's:** Google, AWS, iPhone, Facebook

**2010's:** Cloud-native, NoSQL, mobile-first, SaaS

**2020's:** AI/ML, serverless, edge computing, blockchain (lol)

**Every decade, last decade's hot shit becomes legacy.**

**Oracle in 1995:** The future of enterprise computing

**Oracle in 2025:** Legacy system nobody wants to touch

**AWS in 2015:** The future of cloud computing

**AWS in 2035:** ???

**The cycle never stops.**

Larry's Lesson

**Larry Ellison built Oracle into a $200B company.**

**He's worth $150B personally.**

**He owns Hawaiian islands and America's Cup yachts.**

**But his technology is obsolete.**

**The lesson:**

**You can WIN the hot shit tech race and still become irrelevant.**

**Oracle won the 90's. AWS is winning the 2020's.**

**But both will be replaced.**

**Because hot shit tech always dies.**

**Quality work (like O'Toole's Pulitzer, Commodore 64 games, Walter White's 99.1% purity) survives decades.**

**But platforms, databases, and "essential" tech? Graveyards are full of them.**

---

**P.S.** - This is Post 18. Larry Ellison is worth $150B and still lost. Oracle is 2% cloud market share. The hot shit tech graveyard welcomes all visitors. 🪦

---

**P.P.S.** - Watch: [*Breaking Bad* - Walter's lab evolution](https://www.netflix.com/title/70143836). Equipment changes. Quality obsession doesn't. Same lesson applies to tech. [Netflix](https://www.netflix.com/title/70143836) | [Amazon](https://www.amazon.com/Breaking-Bad-Complete-Bryan-Cranston/dp/B08FQ6N5B6)

---

**P.P.P.S.** - **Graveyard roll call:** Netscape (dead), RealPlayer (forgotten), Yahoo ($125B → $5B), Flash (killed), GeoCities (deleted), AltaVista (closed), AOL ($160B → $4.4B), Oracle (2% cloud share). All were "essential" once. None matter now. Your "hot shit tech" stack will join them. 🪦

---

**P.P.P.P.S.** - DugganUSA uses zero relational databases not because we hate Oracle, but because we refuse to bet on ANY hot shit tech. Flat files + Redis = boring, portable, cheap. When the next hot shit cycle comes, we'll migrate in 35 minutes. Larry can't. That's the difference. 💎

---

**P.P.P.P.P.S.** - Vince Gilligan's genius: Show us Walter's equipment evolve while his purity obsession stays constant. The TOOLS changed. The CRAFT didn't. Oracle's tools are obsolete. But ACID transactions, data integrity, reliability? Those principles survive. Hot shit tech dies. Quality principles are timeless. 🔬