| 3% | “three percent” | Levie: AI slop in legal contracts creates risks code slop never introduced | Aaron Levie | Mar 5, 2026 | — |
| 20× | “20 times” | Levie: AI slop in legal contracts creates risks code slop never introduced | Aaron Levie | Mar 5, 2026 | — |
| 100× | “hundred times” | Levie: AI agents will generate 10 to 100 times more code | Aaron Levie | Mar 5, 2026 | — |
| 100% | “hundred percent” | Becker: Claude Opus 4 Solves Atomic Software Tasks 100% Reliably | Joel Becker | Feb 27, 2026 | Partly supported |
| 4% | “four percent” | Patel: Claude Code's share of GitHub commits doubled to 4% in January | Dylan Patel | Feb 26, 2026 | Supported |
| 2% | “two percent” | Patel: Claude Code's share of GitHub commits doubled to 4% in January | Dylan Patel | Feb 26, 2026 | Supported |
| 80B | “eighty billion” | Patel: Google and Amazon stocks dropped 10% on combined $380B CapEx | Dylan Patel | Feb 26, 2026 | Supported |
| 200B | “two hundred billion” | Patel: Google and Amazon stocks dropped 10% on combined $380B CapEx | Dylan Patel | Feb 26, 2026 | Supported |
| 10% | “10%” | Patel: Google and Amazon stocks dropped 10% on combined $380B CapEx | Dylan Patel | Feb 26, 2026 | Supported |
| $2B | “two billion dollars” | Patel: Anthropic added $2B in monthly revenue at positive margins | Dylan Patel | Feb 26, 2026 | — |
| $40B | “forty billion dollars” | Patel: Meta is taking on $40B in debt for its Louisiana AI cluster | Dylan Patel | Feb 26, 2026 | Partly supported |
| 6.2B | “6.2 billion” | Welling: Billion-Dollar Funding Rounds Signal Exploding AI for Science Bubble | Max Welling | Feb 25, 2026 | — |
| 50% | “50%” | Welling: Perovskite-Silicon Solar Panels Can Theoretically Reach 50% Light Capture | Max Welling | Feb 25, 2026 | Supported |
| 130M | “a hundred thirty million” | Welling: CuspAI has raised $130 million and expanded to 40 people | Max Welling | Feb 25, 2026 | — |
| 5% | “five percent” | O'Laughlin: Being 5% more accurate on EPS models never drives good investment decisions | Doug O'Laughlin | Feb 24, 2026 | — |
| 4% | “four percent” | O'Laughlin: Claude Code captured 4% of GitHub commits in two weeks | Doug O'Laughlin | Feb 24, 2026 | Contradicted |
| 90% | “90%” | O'Laughlin: Reviewing AI agents requires past hands-on manual domain experience | Doug O'Laughlin | Feb 24, 2026 | — |
| 10% | “10%” | O'Laughlin: Reviewing AI agents requires past hands-on manual domain experience | Doug O'Laughlin | Feb 24, 2026 | — |
| 2% | “two percent” | Swyx: Project Stargate represents 2% of US GDP | Shawn Wang | Feb 24, 2026 | Supported |
| 1B | “a billion” | Swyx: GitHub Copilot Has at Least $1 Billion in ARR | Shawn Wang | Feb 24, 2026 | Not publicly verifiable |
| 1T | “a trillion” | O'Laughlin: Google TPU business could be worth $1 trillion | Doug O'Laughlin | Feb 24, 2026 | — |
| 30% | “30%” | O'Laughlin: Google TPU business could be worth $1 trillion | Doug O'Laughlin | Feb 24, 2026 | — |
| 4× | “four X” | O'Laughlin: Manufacturing one bit of HBM takes 4x the DRAM wafer capacity | Doug O'Laughlin | Feb 24, 2026 | Supported |
| 100% | “hundred percent” | O'Laughlin: DRAM prices could rise another 100%, causing demand destruction | Doug O'Laughlin | Feb 24, 2026 | — |
| 100M | “a hundred million” | Swyx: AI context windows will remain capped near 1M for 5-10 years | Shawn Wang | Feb 24, 2026 | Open |
| 1T | “a trillion” | Swyx: AI context windows will remain capped near 1M for 5-10 years | Shawn Wang | Feb 24, 2026 | Open |
| 90% | “90%” | Watkins: About 90% of SWE-bench Verified tasks take under an hour | Olivia Watkins | Feb 23, 2026 | Supported |
| 80% | “80%” | Glaese: OpenAI no longer trusts further score improvements on SWE-bench Verified | Mia Glaese | Feb 23, 2026 | — |
| 61M | “a sixty million” | Casado: AI venture rounds now routinely require complex strategic compute contracts | Martin Casado | Feb 19, 2026 | — |
| 3× | “three times” | Casado: Frontier AI Labs Could Outspend and Consume Application Layer Startups | Martin Casado | Feb 19, 2026 | — |
| 1B | “a billion” | Casado: A $1B model training run economically justifies a custom ASIC | Martin Casado | Feb 19, 2026 | — |
| 2B | “a one billion” | Casado: A $1B model training run economically justifies a custom ASIC | Martin Casado | Feb 19, 2026 | — |
| $200M | “two hundred million dollars” | Casado: A $1B model training run economically justifies a custom ASIC | Martin Casado | Feb 19, 2026 | — |
| 20% | “20%” | Casado: A $1B model training run economically justifies a custom ASIC | Martin Casado | Feb 19, 2026 | — |
| 80% | “80%” | Wang: Compute accounts for roughly 80% of foundation model fundraises | Sarah Wang | Feb 19, 2026 | — |
| 300M | “three hundred million” | Jeff Dean: Distillation originated to compress 50-model ensembles into serviceable form | Jeff Dean | Feb 12, 2026 | Partly supported |
| 95% | “95%” | Dean: AI benchmarks above 95% accuracy offer diminishing returns due to data leakage | Jeff Dean | Feb 12, 2026 | — |
| 1M | “a million” | Dean: Scaling quadratic attention cannot reach billion- or trillion-token context windows | Jeff Dean | Feb 12, 2026 | — |
| 1T | “a trillion” | Dean: Scaling quadratic attention cannot reach billion- or trillion-token context windows | Jeff Dean | Feb 12, 2026 | — |
| 1B | “a billion” | Dean: Scaling quadratic attention cannot reach billion- or trillion-token context windows | Jeff Dean | Feb 12, 2026 | — |
| 30M | “Thirty million” | Jeff Dean: LLM search funnels trillions of tokens down to 100 documents | Jeff Dean | Feb 12, 2026 | — |
| 100× | “hundred X” | Jeff Dean: Design systems to scale 5x to 10x, not 100x | Jeff Dean | Feb 12, 2026 | — |
| 1K | “a thousand” | Dean: Accelerator batching is driven by 1000x SRAM data movement energy costs | Jeff Dean | Feb 12, 2026 | — |
| 10× | “10 X” | Dean: Sparse models offer 10x compute cost efficiency over dense models | Jeff Dean | Feb 12, 2026 | Supported |
| 10× | “10 X” | Dean: Transformers delivered 10x to 100x compute efficiency over LSTMs | Jeff Dean | Feb 12, 2026 | Supported |
| 100× | “hundred X” | Dean: Transformers delivered 10x to 100x compute efficiency over LSTMs | Jeff Dean | Feb 12, 2026 | Supported |
| 2B | “two billion” | Dean: Early 2B-parameter Google Brain model cut ImageNet-22K error by 70% | Jeff Dean | Feb 12, 2026 | Partly supported |
| 70% | “70%” | Dean: Early 2B-parameter Google Brain model cut ImageNet-22K error by 70% | Jeff Dean | Feb 12, 2026 | Partly supported |
| 20× | “20 x” | Dean: AI systems will achieve 20x to 50x lower latency | Jeff Dean | Feb 12, 2026 | Open |
| 50× | “50 x” | Dean: AI systems will achieve 20x to 50x lower latency | Jeff Dean | Feb 12, 2026 | Open |
| 1B | “a billion” | AlphaFold Models Have Fewer Parameters but Higher Compute Costs Than LLMs | Gabriele Corso | Feb 12, 2026 | Supported |
| 2.7× | “2.7 X” | Datology BeyondWeb 3B matches NVIDIA Nemotron 8B in 2.7x less training time | Pratyush Maini | Feb 10, 2026 | Partly supported |
| 2T | “a one trillion” | Bissell: Goodfire performs activation steering on 1-trillion parameter Kimi K2 | Mark Bissell | Feb 5, 2026 | Supported |
| 10× | “10 times” | Garg: Context graph stack will eclipse modern data stack value by 10x | Ashu Garg | Feb 4, 2026 | — |
| 1% | “one percent” | White: Existing LLMs Are Already Capable of Automating Much of Science | Andrew White | Jan 28, 2026 | — |
| 1K | “a thousand” | White: Simulating biological systems hits fundamental limits; empirical lab measurement is unavoidable | Andrew White | Jan 28, 2026 | — |
| 60% | “60%” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 70% | “70%” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 100% | “hundred percent” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 50% | “50%” | White: Cosmos's 50% benchmark score stems from subjective result interpretation | Andrew White | Jan 28, 2026 | Supported |
| 20% | “20%” | Anderson: 20% of pre-ChatGPT global compute went to simulating water | Brandon Anderson | Jan 28, 2026 | — |
| $1M | “a million dollars” | Andrew White: Funding $1M+ Cash Salaries With VC Equity Is Insane | Andrew White | Jan 28, 2026 | — |
| 100× | “hundred X” | White: Future scientists will act as AI agent wranglers exploring 100 ideas simultaneously | Andrew White | Jan 28, 2026 | — |
| 60% | “60%” | Reggio: Brex Card Product Drives 60% of Direct Revenue | James Reggio | Jan 17, 2026 | — |
| 10× | “10 X” | Reggio: Brex structures its AI strategy into corporate, operational, and product pillars | James Reggio | Jan 17, 2026 | — |
| 80% | “80%” | Reggio: Brex targets 80% automated application decisions within 60 seconds | James Reggio | Jan 17, 2026 | — |
| $1M | “a million dollars” | Reggio: Brex sets minimum ICP at $1M revenue or $10K monthly card spend | James Reggio | Jan 17, 2026 | — |
| 100% | “hundred percent” | Brex will freeze engineering headcount at 300 while targeting doubled efficiency | James Reggio | Jan 17, 2026 | — |
| 10T | “10 trillion” | Hill-Smith estimates Gemini 3 Pro parameter count hits 5 to 10 trillion | Micah Hill-Smith | Jan 9, 2026 | — |
| 100× | “hundred times” | GPT-4-level intelligence is now over 100 times cheaper than at launch | Micah Hill-Smith | Jan 9, 2026 | Supported |
| 10× | “10 times” | Reasoning models consume 10x more tokens on average than non-reasoning models | Micah Hill-Smith | Jan 9, 2026 | Supported |
| $1B | “one billion dollars” | Frontier AI research requires up to $100B, far exceeding NSF budgets | Andy Konwinski | Dec 31, 2025 | — |
| $100B | “a hundred billion dollars” | Frontier AI research requires up to $100B, far exceeding NSF budgets | Andy Konwinski | Dec 31, 2025 | — |
| 500× | “500 times” | Bissell: Probing internal model features matches LLM-as-a-judge quality at 500x lower cost | Mark Bissell | Dec 31, 2025 | — |
| 1K | “a thousand” | Wang: 64 layers saturate performance in most reinforcement learning tasks | Kevin Wang | Dec 31, 2025 | Supported |
| 50M | “fifty million” | Arena processes tens of millions of conversations monthly, totaling 250 million | Anastasios Angelopoulos | Dec 31, 2025 | — |
| 25% | “25%” | Angelopoulos: 25% of LMArena platform users write software for a living | Anastasios Angelopoulos | Dec 31, 2025 | — |
| 1M | “a million” | Gradio scaled Arena to 1 million monthly active users before migration | Anastasios Angelopoulos | Dec 31, 2025 | Supported |
| 9% | “nine percent” | Arena sampled open-source models at 60/40, debunking Leaderboard Illusion paper | Anastasios Angelopoulos | Dec 31, 2025 | Supported |
| 60% | “60%” | Arena sampled open-source models at 60/40, debunking Leaderboard Illusion paper | Anastasios Angelopoulos | Dec 31, 2025 | Supported |
| 1T | “a trillion” | Nair: LLM agents will hit $1T before robotics hits $10B | Ashvin Nair | Dec 30, 2025 | Didn’t hold up |
| 11B | “a ten billion” | Nair: LLM agents will hit $1T before robotics hits $10B | Ashvin Nair | Dec 30, 2025 | Didn’t hold up |
| 20% | “20%” | Nair: OpenAI internal models surpassed forecasters' 2027 benchmark targets before o1 launch | Ashvin Nair | Dec 30, 2025 | — |
| 1T | “a trillion” | Nair: Continual learning during deployment does not risk model capacity overload | Ashvin Nair | Dec 30, 2025 | — |
| $100M | “a hundred million dollars” | Catanzaro: $100M+ AI seed rounds without roadmaps happen frequently | Sarah Catanzaro | Dec 30, 2025 | — |
| $2B | “two billion dollars” | Linux Foundation avoided frontier LLMs due to multibillion-dollar compute and data costs | Jim Zemlin | Dec 28, 2025 | — |
| 10× | “10 times” | Zemlin: MCP and AAIF adoption will move ten times faster than Kubernetes | Jim Zemlin | Dec 28, 2025 | — |
| 10× | “10 X” | Yegge: OpenAI sees a 10x productivity gap between AI adopters and non-adopters | Steve Yegge | Dec 26, 2025 | — |
| 10× | “10 times” | Yegge: 10x AI productivity turns merging into architectural rewrites | Steve Yegge | Dec 26, 2025 | — |
| 50% | “50%” | Chen: Approximately 50% of OpenAI employees adopted Codex at launch | Bill Chen | Dec 26, 2025 | — |
| 6M | “six million” | Nelson: Roboflow Records 106 Million Annotations Powered by Meta SAM | Joseph Nelson | Dec 18, 2025 | — |
| 8M | “eight million” | Nelson: Users Ran 8 Million Inferences in SAM 3's First Five Days | Joseph Nelson | Dec 18, 2025 | — |
| 70% | “70%” | Ravi: Over 70% of SAM 3 Dataset Annotations Are Negative Phrases | Nikhila Ravi | Dec 18, 2025 | Supported |
| 99% | “99%” | Pliny: Crafting AI jailbreak prompts is 99% intuition over technical knowledge | Pliny the Liberator | Dec 16, 2025 | — |
| 50M | “fifty million” | Mirzadegan: Hiring $50M+ ARR Sales Reps at Seed/Series A Usually Fails | Joubin Mirzadegan | Dec 12, 2025 | — |
| 95% | “95%” | Mirzadegan: Salesforce's new CPQ replacement product is non-existent or flimsy | Joubin Mirzadegan | Dec 12, 2025 | — |
| 1M | “a million” | Superhuman operates with 50 engineers and roughly 100,000 paying users | Loïc Houssier | Dec 11, 2025 | — |
| 80% | “80%” | General Intuition aims to drive 80% of AI physical interactions by 2030 | Pim de Witte | Dec 6, 2025 | — |
| 1K | “a thousand” | Johnson: AI compute per model has scaled one million-fold since 2012 | Justin Johnson | Nov 25, 2025 | Supported |
| 1M | “a million” | Johnson: AI compute per model has scaled one million-fold since 2012 | Justin Johnson | Nov 25, 2025 | Supported |
| 40M | “forty million” | Li: Evolution spent 540M years on spatial intelligence versus 0.5M on language | Fei-Fei Li | Nov 25, 2025 | Supported |
| 1.5M | “half a million” | Li: Evolution spent 540M years on spatial intelligence versus 0.5M on language | Fei-Fei Li | Nov 25, 2025 | Supported |
| $100M | “a hundred million dollars” | Building a custom IDE would cost AI coding startups $100M | Matthias Wagner | Nov 22, 2025 | — |
| 1M | “a million” | Flux boasts a user-generated library of over one million electronic parts | Matthias Wagner | Nov 22, 2025 | Partly supported |
| 10× | “10 times” | Wagner: Task-specific knowledge injection effectively reduces LLM hallucinations | Matthias Wagner | Nov 22, 2025 | — |
| 26× | “26 X” | Flux reached 7,000 paying customers and grew 26x year-over-year | Matthias Wagner | Nov 22, 2025 | — |
| $1M | “a million dollars” | Hezarkhani: Tenex will have engineers making over $1M cash next year | Arman Hezarkhani | Nov 19, 2025 | — |
| 100% | “hundred percent” | Lieberman: Tenex's business growth is 100% human talent constrained, not agent constrained | Alex Lieberman | Nov 19, 2025 | — |
| 100M | “hundred million” | Glean Operates at a Several Hundred Million Dollar Revenue Scale | Deedy Das | Nov 14, 2025 | Supported |
| 80% | “80%” | Anthropic Maintains an 80% One-Year Employee Retention Rate | Deedy Das | Nov 14, 2025 | Supported |
| 80% | “80%” | Voice Dictation Startup Wispr Achieves a Zero-Edit Rate Above 80% | Deedy Das | Nov 14, 2025 | — |
| 90% | “90%” | Diffusion Models Achieve 80% of Autoregressive Quality at One-Tenth the Cost | Deedy Das | Nov 14, 2025 | — |
| $7B | “seven billion dollars” | OpenAI Spent $7 Billion on Compute, With $5 Billion for R&D | Shawn Wang | Nov 14, 2025 | Contradicted |
| 800M | “eight hundred million” | OpenAI Spent $7 Billion on Compute, With $5 Billion for R&D | Shawn Wang | Nov 14, 2025 | Contradicted |
| 2B | “two billion” | OpenAI Spent $7 Billion on Compute, With $5 Billion for R&D | Shawn Wang | Nov 14, 2025 | Contradicted |
| 1M | “a million” | Vercel's v0 took nine months to reach $1M ARR | Jared Palmer | Nov 10, 2025 | — |
| 90% | “90%” | HackerRank trained model detecting external coding tools with 85-90% precision | Vivek Ravisankar | Nov 8, 2025 | — |
| 25% | “25%” | HackerRank: Junior developer assessments and interviews are up 20-25% year-over-year | Vivek Ravisankar | Nov 8, 2025 | — |
| 1M | “a million” | Ravisankar is working with Rwanda to train one million African coders | Vivek Ravisankar | Nov 8, 2025 | — |
| 25M | “twenty-five million” | Chan: CELLxGENE corpus reached 125M cells with 75% community contribution | Priscilla Chan | Nov 6, 2025 | Supported |
| 25% | “25%” | Chan: CELLxGENE corpus reached 125M cells with 75% community contribution | Priscilla Chan | Nov 6, 2025 | Supported |
| 75% | “75%” | Chan: CELLxGENE corpus reached 125M cells with 75% community contribution | Priscilla Chan | Nov 6, 2025 | Supported |
| 300× | “300 X” | AMD MI300X outperforms Nvidia H100 on FlashAttention-2 and memory-bound workloads | Quentin Anthony | Nov 3, 2025 | Supported |
| 300× | “300 X” | AMD MI300X GEMM performance increased from 400 to 650 TFLOPS via software | Quentin Anthony | Nov 3, 2025 | Supported |
| 100× | “hundred times” | Rumbelow: Leap Labs is roughly 100x faster than manual analysis | Jessica Rumbelow | Nov 2, 2025 | — |
| $1.4T | “1.4 trillion dollars” | Sands: Stripe processes $1.4 trillion annually, roughly 1.3% of global GDP | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 1.3% | “1.3%” | Sands: Stripe processes $1.4 trillion annually, roughly 1.3% of global GDP | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 47% | “47%” | Sands: Roughly 47% of payments leaders cite friendly fraud as biggest issue | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 200M | “two hundred million” | Sands: Stripe Link Surpasses 200 Million Consumers | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 58% | “58%” | Sands: 58% of Lovable's Payment Volume Flows Through Stripe Link | Emily Glassberg Sands | Oct 30, 2025 | — |
| 1M | “one million” | Sands: Over 1M Shopify merchants and Salesforce joining ChatGPT checkout | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 7× | “seven times” | Sands: Stripe Merchants Grew Seven Times Faster Than S&P 500 | Emily Glassberg Sands | Oct 30, 2025 | Supported |
| 70% | “70%” | Sands: 65% to 70% of Stripe engineers use AI coding assistants daily | Emily Glassberg Sands | Oct 30, 2025 | — |
| 50× | “50 times” | Sands: Manual writing forces first-principles reasoning that LLMs dangerously bypass | Emily Glassberg Sands | Oct 30, 2025 | — |
| $3M | “three million dollars” | Sands: Switching to o3-mini saved Stripe $3M yearly on risk task | Emily Glassberg Sands | Oct 30, 2025 | — |
| 1M | “A million” | Sands: Top 100 AI startups reach ARR milestones 2-3x faster than SaaS | Emily Glassberg Sands | Oct 30, 2025 | — |
| 10M | “ten million” | Sands: Top 100 AI startups reach ARR milestones 2-3x faster than SaaS | Emily Glassberg Sands | Oct 30, 2025 | — |
| 30M | “thirty million” | Sands: Top 100 AI startups reach ARR milestones 2-3x faster than SaaS | Emily Glassberg Sands | Oct 30, 2025 | — |
| 3× | “three times” | Sands: Top 100 AI startups reach ARR milestones 2-3x faster than SaaS | Emily Glassberg Sands | Oct 30, 2025 | — |
| 10% | “10%” | Webster: Over 10% of Fortune 500 companies use Promptfoo | Ian Webster | Oct 24, 2025 | — |
| 20% | “20%” | Webster: Finding enterprise AI vulnerabilities is only 20% of the challenge | Ian Webster | Oct 24, 2025 | — |
| 95% | “95%” | Webster: AI guardrails are a commodity and easy to build | Ian Webster | Oct 24, 2025 | — |
| 10T | “10 trillion” | Swix: Frontier model sizes have likely peaked around 2 trillion parameters | Shawn Wang | Oct 20, 2025 | — |
| 6.5T | “6.5 trillion” | Bakouch: SmolLM 2 scored random on MMLU until 6.5 trillion tokens | Elie Bakouch | Oct 20, 2025 | Supported |
| 6T | “six trillion” | Bakouch: SmolLM 2 scored random on MMLU until 6.5 trillion tokens | Elie Bakouch | Oct 20, 2025 | Supported |
| 70% | “70%” | Merrill: AI evals will shift to observing real jobs over 2-3 years | Mike Merrill | Oct 18, 2025 | — |
| 70% | “70%” | Merrill: 60% to 70% of SWE-bench Verified tasks come from Django | Mike Merrill | Oct 18, 2025 | Contradicted |
| 15% | “15%” | Alex Shaw: Agent frameworks can impact benchmark scores by up to 15% | Alex Shaw | Oct 18, 2025 | Supported |
| $1K | “a thousand dollars” | Merrill: AI agents could ace Terminal-Bench with $1,000 compute per task | Mike Merrill | Oct 18, 2025 | — |
| 1M | “a million” | Corbitt: OpenPipe Hit $1M ARR Within Eight Months of Launch | Kyle Corbitt | Oct 16, 2025 | — |
| 90% | “90%” | Corbitt: Fine-tuning offers poor ROI for 90% of unconstrained use cases | Kyle Corbitt | Oct 16, 2025 | — |
| 60% | “60%” | Corbitt: 55-60% chance RL becomes the standard pattern for deploying scale agents | Kyle Corbitt | Oct 16, 2025 | — |
| 50% | “50%” | Corbitt: Prompt optimization methods like JEPA failed OpenPipe's agent benchmarks | Kyle Corbitt | Oct 16, 2025 | — |
| 50% | “50%” | Swix: Open-source token generation share will rise but stay well below 50% | Shawn Wang | Oct 16, 2025 | — |
| 20% | “20%” | Alessio: Open-source token share could reach 15-20% excluding coding | Alessio Fanelli | Oct 16, 2025 | — |
| 10× | “10 times” | Corbitt: AI inference could be 10x larger if reliability issues are solved | Kyle Corbitt | Oct 16, 2025 | — |
| 4M | “four million” | OpenAI reports 4 million active developers at DevDay 2025 | Shawn Wang | Oct 7, 2025 | Supported |
| 10% | “10%” | AI industry has completed only 10% of necessary agent evaluation progress | Sherwin Wu | Oct 7, 2025 | — |
| 40% | “40%” | OpenAI Codex successfully one-shots entire features 30% to 40% of the time | Sherwin Wu | Oct 7, 2025 | — |
| 20× | “20 times” | Agarwal: Sequoia met 20 AI SRE startups before backing Traversal | Anish Agarwal | Oct 5, 2025 | — |
| 40% | “40%” | Dwivedi: AI self-healing for complex incidents is 6-12 months away | Raaz Dwivedi | Oct 5, 2025 | — |
| 85% | “85%” | Field: Deep learning was necessary to achieve 100% in visual generation | Dylan Field | Oct 2, 2025 | — |
| 100% | “hundred percent” | Field: Deep learning was necessary to achieve 100% in visual generation | Dylan Field | Oct 2, 2025 | — |
| 2.1B | “a 1.1 billion” | Feldman: Cerebras raised $1.1B at an $8.1B valuation | Andrew Feldman | Oct 1, 2025 | Supported |
| 9.1B | “an 8.1 billion” | Feldman: Cerebras raised $1.1B at an $8.1B valuation | Andrew Feldman | Oct 1, 2025 | Supported |
| 20× | “20 times” | Feldman: Cerebras is 20 times faster than Nvidia B200 GPUs | Andrew Feldman | Oct 1, 2025 | Contradicted |
| 2,625× | “2625 times” | Feldman: Cerebras provides 2,625x more memory bandwidth than traditional GPUs | Andrew Feldman | Oct 1, 2025 | Supported |
| 30B | “thirty billion” | Enterprises retrain 10B-30B open models from scratch for legal data compliance | Andrew Feldman | Oct 1, 2025 | — |
| 25× | “25 times” | Feldman: Non-compute bottlenecks masked Cerebras' 20-25x speedup for a hyperscaler | Andrew Feldman | Oct 1, 2025 | — |
| 50% | “50%” | Slack: Amp is growing more than 50% month-over-month | Quinn Slack | Sep 25, 2025 | — |
| 100× | “hundred times” | Slack: Asynchronous background agents will dominate AI inference and output | Quinn Slack | Sep 25, 2025 | — |
| 15× | “15 times” | Ball: Amp core team has 8 people, ships 15 times daily without code reviews | Thorsten Ball | Sep 25, 2025 | — |
| 200M | “Two hundred million” | Fanelli: Cursor Default Model Switch Cost Anthropic $200M in ARR | Alessio Fanelli | Sep 25, 2025 | — |
| 98% | “98%” | Ball: Complex sub-agent workflows will result in user hangovers | Thorsten Ball | Sep 25, 2025 | — |
| 100% | “hundred percent” | Slack: Competitors discounted products up to 100% to win deals against Amp | Quinn Slack | Sep 25, 2025 | — |
| 85% | “85%” | Slack: Competitors discounted products up to 100% to win deals against Amp | Quinn Slack | Sep 25, 2025 | — |
| 95% | “95%” | Slack: Amp only runs tests correctly 90-95% of the time | Quinn Slack | Sep 25, 2025 | — |
| 20M | “twenty million” | Rajpal: Waymo had 20 million real-world miles versus 20 billion in simulation | Shreya Rajpal | Sep 25, 2025 | Supported |
| 20B | “twenty billion” | Rajpal: Waymo had 20 million real-world miles versus 20 billion in simulation | Shreya Rajpal | Sep 25, 2025 | Supported |
| 10× | “10 X” | Bachman: Power Retention Delivers 100x Inference Speedup at 64k Context | Diego Bachman | Sep 23, 2025 | Open |
| 100× | “hundred X” | Bachman: Power Retention Delivers 100x Inference Speedup at 64k Context | Diego Bachman | Sep 23, 2025 | Open |
| 35% | “35%” | Bachman: PowerCoder-3B reaches 35% HumanEval accuracy versus StarCoder's 30% | Diego Bachman | Sep 23, 2025 | Open |
| 30% | “30%” | Bachman: PowerCoder-3B reaches 35% HumanEval accuracy versus StarCoder's 30% | Diego Bachman | Sep 23, 2025 | Open |
| 4B | “a three billion” | Bachman: Manifest AI built Manifesto to repair repositories using PowerCoder-3B | Diego Bachman | Sep 23, 2025 | — |
| 30B | “thirty billion” | Bachman: Manifest plans 30-billion-parameter foundation model using Power Retention | Diego Bachman | Sep 23, 2025 | — |
| 90% | “90%” | Bachman: 90% of documents in web pre-training datasets are under 2k context | Diego Bachman | Sep 23, 2025 | Supported |
| 10M | “ten million” | Falk: Zeus originally targeted Global 2000 companies for $5M-$10M contracts | Brendan Falk | Sep 18, 2025 | — |
| 1M | “a million” | Falk: Scoping large custom enterprise contracts requires only marginally more work than small ones | Brendan Falk | Sep 18, 2025 | — |
| 6M | “a five million” | Falk: Scoping large custom enterprise contracts requires only marginally more work than small ones | Brendan Falk | Sep 18, 2025 | — |
| $10M | “ten million dollars” | Falk: Enterprise AI services only work at $5M-$10M scale or single-vertical focus | Brendan Falk | Sep 18, 2025 | — |
| $5M | “five million dollars” | Falk: Enterprise AI services only work at $5M-$10M scale or single-vertical focus | Brendan Falk | Sep 18, 2025 | — |
| 95% | “95%” | Martin: Claude Code triggers context compaction at 95% of context window | Lance Martin | Sep 11, 2025 | Supported |
| 2M | “two million” | Yurtseven: Fal.ai has around two million developers on its platform | Gorkem Yurtseven | Sep 8, 2025 | — |
| 100M | “a hundred million” | Yurtseven confirms Fal.ai has surpassed $100M in revenue | Gorkem Yurtseven | Sep 8, 2025 | — |
| 2M | “two million” | Taskaya: Flux grew Fal's revenue from $2M to $20M in two months | Batuhan Taskaya | Sep 8, 2025 | — |
| 10M | “ten million” | Taskaya: Flux grew Fal's revenue from $2M to $20M in two months | Batuhan Taskaya | Sep 8, 2025 | — |
| 16% | “16%” | Taskaya: Custom ASICs do not make sense given low Nvidia GEMM overhead | Batuhan Taskaya | Sep 8, 2025 | — |
| 50% | “50%” | Yurtseven: Video models now account for over 50% of Fal's revenue | Gorkem Yurtseven | Sep 8, 2025 | — |
| 100% | “hundred percent” | Yurtseven: Video models now account for over 50% of Fal's revenue | Gorkem Yurtseven | Sep 8, 2025 | — |
| 1% | “one percent” | Taskaya: NSFW content makes up less than 1% of Fal traffic | Batuhan Taskaya | Sep 8, 2025 | — |
LATENT SPACELEDGER All shows Try the podcast clipper →