| 100% | “hundred percent” | Burazin: CPU environments must spin up instantly to prevent costly GPU idle time | Ivan Burazin | May 21, 2026 | — |
| 100% | “hundred percent” | Ethan He: Video models require image foundations and 100% synthetic caption pairs | Ethan He | Jun 1, 2026 | — |
| 100% | “hundred percent” | Parakhin: Daily AI tool adoption at Shopify approaches 100 percent | Mikhail Parakhin | Apr 22, 2026 | — |
| 100% | “hundred percent” | Cooper: The vast majority of Railway workloads run on bare metal | Jake Cooper | May 20, 2026 | — |
| 100% | “100%” | Andreessen: AI will never transform existing US K-12 public classrooms | Marc Andreessen | Apr 3, 2026 | Didn’t hold up |
| 100% | “hundred percent” | Rieseberg: Sandboxing bypasses need for 100% model alignment | Felix Rieseberg | Mar 17, 2026 | — |
| 100% | “hundred percent” | Shah: LoCoMo benchmark is flawed because models pass via context dumping | Dhravya Shah | Mar 9, 2026 | — |
| 100% | “hundred percent” | Becker: Claude Opus 4 Solves Atomic Software Tasks 100% Reliably | Joel Becker | Feb 27, 2026 | Partly supported |
| 100% | “hundred percent” | O'Laughlin: DRAM prices could rise another 100%, causing demand destruction | Doug O'Laughlin | Feb 24, 2026 | — |
| 100% | “hundred percent” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 100% | “hundred percent” | Brex will freeze engineering headcount at 300 while targeting doubled efficiency | James Reggio | Jan 17, 2026 | — |
| 100% | “hundred percent” | Lieberman: Tenex's business growth is 100% human talent constrained, not agent constrained | Alex Lieberman | Nov 19, 2025 | — |
| 100% | “hundred percent” | Field: Deep learning was necessary to achieve 100% in visual generation | Dylan Field | Oct 2, 2025 | — |
| 100% | “hundred percent” | Slack: Competitors discounted products up to 100% to win deals against Amp | Quinn Slack | Sep 25, 2025 | — |
| 100% | “hundred percent” | Yurtseven: Video models now account for over 50% of Fal's revenue | Gorkem Yurtseven | Sep 8, 2025 | — |
| 100% | “100%” | CloudChef Operates with 100% Decision Autonomy and 90% Action Autonomy | Nikhil Abraham | May 31, 2025 | — |
| 100% | “hundred percent” | Brown: Prompting alone cannot reliably force LLMs to use thinking tokens | Will Brown | May 23, 2025 | — |
| 100% | “hundred percent” | Optimal future AI coding interfaces will not evolve from traditional IDEs | Matan Grinberg | May 29, 2025 | — |
| 100% | “hundred percent” | Up to 75% of Future UX Interfaces Will Be Voice-Driven | Kwindla Hultman Kramer | May 6, 2025 | — |
| 100% | “hundred percent” | Conrad: Spot GPU cluster utilization nears 100% through dynamic price clearing | Evan Conrad | Apr 11, 2025 | — |
| 100% | “hundred percent” | Klein: Web scraping workflows should use a tiered waterfall architecture | Paul Klein | Feb 28, 2025 | — |
| 100% | “hundred percent” | McAteer: o1 is the first model to achieve one-shot codebase implementation | Dan McAteer | Jan 17, 2025 | — |
| 100% | “hundred percent” | Beauchamp: Small trading funds hold an advantage exploiting niche market anomalies | William Beauchamp | Jan 26, 2025 | — |
| 100% | “hundred percent” | Ben Allal: LLMs can be trained with entirely synthetic pipelines | Loubna Ben Allal | Dec 24, 2024 | Partly supported |
| 100% | “hundred percent” | Goyal: Braintrust saw nearly 100% OpenAI market share pre-Claude 3 | Ankur Goyal | Oct 11, 2024 | — |
| 100% | “hundred percent” | Hu: AI Models Should Eventually Reach 100% on SWE-bench Verified | Jesse Hu | Oct 19, 2024 | Open |
| 100% | “100%” | Jamil: OpenAI and Cohere Overlap Prefill and Generation to Maximize GPU Utilization | Umar Jamil | Sep 19, 2024 | — |
| 100% | “hundred percent” | Carlini: Anyone claiming 0% or 100% certainty on 5-year AI capabilities is probably wrong | Nicholas Carlini | Aug 28, 2024 | — |
| 100% | “hundred percent” | Function calling benchmarks like BFCL are largely saturated | Michelle Pokrass | Sep 17, 2024 | — |
| 100% | “hundred percent” | Multi-step agentic apps fail at 95% reliability due to compounded errors | Michelle Pokrass | Sep 17, 2024 | — |
| 100% | “hundred percent” | Albrecht: Coding agents communicating uncertainty are far more useful than slightly more accurate ones | Josh Albrecht | Jun 25, 2024 | — |
| 100% | “100%” | Conover: AI companies hire systems engineers, traditional software hires AI talent | Mike Conover | Jun 11, 2024 | — |
| 100% | “100%” | Conover: AI model developers are absolutely overfitting to public evaluation benchmarks | Mike Conover | Jun 11, 2024 | — |
| 100% | “hundred percent” | Lambert: RLHF reward models achieve only 65% to 75% validation agreement | Nathan Lambert | Jan 11, 2024 | Supported |
| 100% | “hundred percent” | Sean Lie: 95% of High-Quality Open Models Come From Chinese Labs | Sean Lie | Sep 2, 2026 | — |
| 99% | “99%” | Xin: Unifying storage delivers 99% of HTAP database benefits | Reynold Xin | Jun 24, 2026 | — |
| 99% | “99%” | Hong: Axiom's unmodified Putnam system achieved 99% on Verina benchmark | Carina Hong | Jun 3, 2026 | Open |
| 99% | “99%” | Stay Sassy PM: Leaders know what to do 99% of the time | Stay Sassy PM | Apr 13, 2026 | — |
| 99% | “99%” | Simon Last: 99% of Agent Failures Are Tool Bugs, Not Model Flaws | Simon Last | Apr 15, 2026 | — |
| 99% | “99%” | Horthy: Top 1% of AI builders don't use popular public frameworks | Dex Horthy | Mar 6, 2026 | — |
| 99% | “99%” | Pliny: Crafting AI jailbreak prompts is 99% intuition over technical knowledge | Pliny the Liberator | Dec 16, 2025 | — |
| 99% | “99%” | Hou: 99% of AI editor rules file contents will be automatically inferred | Kevin Hou | Jul 28, 2025 | — |
| 99% | “99%” | Windsurf aims to shift coding workflows from 80% agent to 99% agent | Kevin Hou | Jul 28, 2025 | — |
| 99% | “99%” | 99% of Current Monetizable Voice AI Use Cases Are Telephony | Kwindla Hultman Kramer | May 6, 2025 | — |
| 99% | “99%” | Ben-Smith: Snipd indexes 99% of all podcasts in-house | Kevin Ben-Smith | Mar 14, 2025 | — |
| 99% | “99%” | Neubig: Browsers, Terminals, And Code Editors Constitute The Core Agent Toolset | Graham Neubig | Dec 25, 2024 | — |
| 99% | “99%” | Schluntz: Reliability, not demo capability, is the bottleneck for robotics | Erik Schluntz | Nov 28, 2024 | — |
| 99% | “99%” | Yao: Enterprise customer support AI requires 99% reliability over simple tasks, not search | Shunyu Yao | Sep 27, 2024 | — |
| 99% | “99%” | Howard: Developers should distribute merged adapters rather than merged models | Jeremy Howard | Aug 17, 2024 | — |
| 99% | “99%” | 99% of Data in Multi-Gigabyte Docker Images Is Never Read | Erik Bernhardsson | Feb 19, 2024 | — |
| 98% | “98%” | Beam: CAR-T Therapy Likely Would Have Failed Without Serendipitous Intervention | Andy Beam | Jul 16, 2026 | — |
| 98% | “98%” | Levie: Access governance will decide if enterprise AI agents leak data constantly | Aaron Levie | Mar 5, 2026 | — |
| 98% | “98%” | Ball: Complex sub-agent workflows will result in user hangovers | Thorsten Ball | Sep 25, 2025 | — |
| 97% | “97%” | Fanelli: ExaFunction customer cut compute costs 97% on single GPU | Alessio Fanelli | Jul 28, 2025 | — |
| 96% | “96%” | Krentsel: Exo autonomously re-architected its Discord adapter, cutting costs by 96% | Alex Krentsel | Aug 15, 2026 | Open |
| 96% | “96%” | Most AI clusters fail to hit Google's 96% node utilization standard | Anjney Midha | Jun 18, 2026 | — |
| 95% | “95%” | Most AI clusters fail to hit Google's 96% node utilization standard | Anjney Midha | Jun 18, 2026 | — |
| 95% | “95%” | Yan: Roughly 95% of Devin's memories are auto-generated | Walden Yan | May 28, 2026 | — |
| 95% | “95%” | Yasser Elsaid: 95% of AI customer service limitations stem from harness architecture | Yasser Elsaid | May 2, 2026 | — |
| 95% | “95%” | Younis: Vehicle software testing is 95% CI/CD, 4% rig, 1% physical | Qasar Younis | Apr 27, 2026 | — |
| 95% | “95%” | Eskildsen: Turbopuffer reduced Cursor's infrastructure costs by 95% | Simon Eskildsen | Mar 12, 2026 | — |
| 95% | “95%” | Dean: AI benchmarks above 95% accuracy offer diminishing returns due to data leakage | Jeff Dean | Feb 12, 2026 | — |
| 95% | “95%” | Mirzadegan: Salesforce's new CPQ replacement product is non-existent or flimsy | Joubin Mirzadegan | Dec 12, 2025 | — |
| 95% | “95%” | Webster: AI guardrails are a commodity and easy to build | Ian Webster | Oct 24, 2025 | — |
| 95% | “95%” | Slack: Amp only runs tests correctly 90-95% of the time | Quinn Slack | Sep 25, 2025 | — |
| 95% | “95%” | Martin: Claude Code triggers context compaction at 95% of context window | Lance Martin | Sep 11, 2025 | Supported |
| 95% | “95%” | Ganatra: 95% of Composio integrations are built and maintained by agents | Soham Ganatra | Aug 4, 2025 | — |
| 95% | “95%” | Drew Houston: Regular expressions solve 95% of email triage NLP tasks | Drew Houston | Oct 18, 2024 | — |
| 95% | “95%” | Multi-step agentic apps fail at 95% reliability due to compounded errors | Michelle Pokrass | Sep 17, 2024 | — |
| 95% | “95%” | Hsu: 95% of Retool customers are developers | David Hsu | Feb 7, 2024 | — |
| 95% | “95%” | Sean Lie: 95% of High-Quality Open Models Come From Chinese Labs | Sean Lie | Sep 2, 2026 | — |
| 94% | “94%” | Robicheaux: 2B PaliGemma 2 Beats ChatGPT on MMVP with 47.3% | Peter Robicheaux | Dec 22, 2024 | Supported |
| 93% | “93%” | Sohmers: Positron AI hardware achieves 93% of theoretical memory bandwidth | Thomas Sohmers | Aug 18, 2025 | Open |
| 90% | “90%” | Kant: Foundation model building is 90% engineering rather than research | Eiso Kant | Jul 22, 2026 | — |
| 90% | “90%” | McPartlon: AlphaFold-Multimer gets antibody-antigen prediction right only 11% of the time | Matt McPartlon | Aug 11, 2026 | Supported |
| 90% | “90%” | Awais: AI design slop is a contract gap, not a model capability deficit | Ahmad Awais | Jun 6, 2026 | — |
| 90% | “90%” | Daytona's mean compute utilization is 15%, peaking at 90% | Ivan Burazin | May 21, 2026 | — |
| 90% | “90%” | Lee: Over 90% of healthcare alerts are ignored | Janie Lee | May 14, 2026 | Supported |
| 90% | “90%” | Proper AI harnesses enable 80% to 90% customer issue resolution rates | Yasser Elsaid | May 2, 2026 | — |
| 90% | “90%” | Manning: Vision understanding stalled; language does 90% of work in VLMs | Chris Manning | Apr 2, 2026 | — |
| 90% | “90%” | O'Laughlin: Reviewing AI agents requires past hands-on manual domain experience | Doug O'Laughlin | Feb 24, 2026 | — |
| 90% | “90%” | Watkins: About 90% of SWE-bench Verified tasks take under an hour | Olivia Watkins | Feb 23, 2026 | Supported |
| 90% | “90%” | Diffusion Models Achieve 80% of Autoregressive Quality at One-Tenth the Cost | Deedy Das | Nov 14, 2025 | — |
| 90% | “90%” | HackerRank trained model detecting external coding tools with 85-90% precision | Vivek Ravisankar | Nov 8, 2025 | — |
| 90% | “90%” | Corbitt: Fine-tuning offers poor ROI for 90% of unconstrained use cases | Kyle Corbitt | Oct 16, 2025 | — |
| 90% | “90%” | Bachman: 90% of documents in web pre-training datasets are under 2k context | Diego Bachman | Sep 23, 2025 | Supported |
| 90% | “90%” | Landgraf: Ona has a 90% gross margin on enterprise offering | Johannes (Johannes Landgraf) | Sep 1, 2025 | — |
| 90% | “90%” | Huber: Regex handles 90% of code queries; embeddings add marginal improvement | Jeff Huber | Aug 19, 2025 | — |
| 90% | “90%” | Palazzolo: Meta should focus on app integration over frontier models | Stephanie Palazzolo | Aug 6, 2025 | — |
| 90% | “90%” | Scott Wu: Software engineers spend 80-90% of time on implementation | Scott Wu | Jul 28, 2025 | — |
| 90% | “90%” | Hou: 90% of code written by Windsurf users is generated by Cascade | Kevin Hou | Jul 28, 2025 | — |
| 90% | “90%” | Windsurf aims to shift coding workflows from 80% agent to 99% agent | Kevin Hou | Jul 28, 2025 | — |
| 90% | “90%” | Hsu: 80% to 90% of Tech for Superhuman AI Tutors Exists | Andrew Hsu | Jul 11, 2025 | — |
| 90% | “90%” | Hsu: Speak has 90% of product team in SF and only hires there | Andrew Hsu | Jul 11, 2025 | — |
| 90% | “90%” | Olivia Moore: 90% of TikTok and Reels feeds are AI-generated video | Olivia Moore | Jul 9, 2025 | — |
| 90% | “90%” | Morris: New embedding inversion model exactly recovers 90% of source text | Jack Morris | Jul 2, 2025 | Supported |
| 90% | “90%” | Alberti: 80% to 90% of DeepWiki users view pre-indexed popular repos | Silas Alberti | May 21, 2025 | — |
| 90% | “90%” | CloudChef Operates with 100% Decision Autonomy and 90% Action Autonomy | Nikhil Abraham | May 31, 2025 | — |
| 90% | “90%” | Agarwal: Synthetic data distillation achieves 80% to 90% of target gains | Rishabh Agarwal | Mar 23, 2025 | — |
| 90% | “90%” | A 90% SWE-Bench Score Can Still Fall Flat in Customer Environments | Misha Laskin | Mar 7, 2025 | — |
| 90% | “90%” | Snipd builds on Python, GCP, and Flutter for cross-platform clients | Kevin Ben-Smith | Mar 14, 2025 | — |
| 90% | “90%” | Alessio Fanelli: GPT-4o Search jumps to 90% accuracy on simple QA | Alessio Fanelli | Mar 11, 2025 | Partly supported |
| 90% | “90%” | Klein: Browserbase delivers 90% of computer-use functionality at 10% of OS cost | Paul Klein | Feb 28, 2025 | — |
| 90% | “90%” | Roucher: AI agents will reach a 90% GAIA score by 2026 | Aymeric (Emmerich) | Feb 13, 2025 | Held up |
| 90% | “90%” | Agarwal: 90% of production LLM use cases do not use automatic routing | Rohit Agarwal | Feb 5, 2025 | — |
| 90% | “90%” | Reddy: OpenAI's share of enterprise LLM spend dropped from 90% to 60% | Pranav Reddy | Dec 21, 2024 | Supported |
| 90% | “90%” | Neubig: Agents Solve 80 To 90 Percent Of Tasks With Feedback | Graham Neubig | Dec 25, 2024 | — |
| 90% | “90%” | Cerebras weight streaming prunes up to 90% of data without accuracy loss | Sarah Chieng | Dec 7, 2024 | Supported |
| 90% | “90%” | Crivello: AI has 10% p(doom) and 90% chance of disease-free utopia | Florent Crivello | Nov 15, 2024 | — |
| 90% | “90%” | Goyal: Simple tool-calling prompts cover 80% to 90% of AI use cases | Ankur Goyal | Oct 11, 2024 | — |
| 90% | “90%” | Houston: Dropbox Went 90% Remote to Act as Distributed Work Lab | Drew Houston | Oct 18, 2024 | — |
| 90% | “90%” | Yao: Reliable tool design accounts for 90% of agent performance | Shunyu Yao | Sep 27, 2024 | — |
| 90% | “90%” | Carlini: 90% of scientific research is routine work that AI can automate | Nicholas Carlini | Aug 28, 2024 | — |
| 90% | “90%” | Schulhoff: Few-Shot Exemplar Order Can Shift Model Accuracy From 0% to 90% | Sander Schulhoff | Sep 20, 2024 | Supported |
| 90% | “90%” | Liu: 90% of users ask how to improve LLM app performance | Jerry Liu | Oct 12, 2023 | — |
| 90% | “90%” | Hotz: Tinybox will be 5x faster than H100 systems per dollar | George Hotz | Jun 20, 2023 | — |
| 90% | “90%” | Slack says hiring early PMs or marketers is an outdated practice | Quinn Slack | Sep 7, 2026 | — |
| 88% | “88%” | Polu: Dust averages 60% to 70% weekly active penetration in enterprise accounts | Stanislas Polu | Nov 11, 2024 | — |
| 85% | “85%” | Park: Generative agent digital twins replicate human behavior at 85% accuracy | Joon Sung Park | Aug 21, 2026 | Supported |
| 85% | “85%” | Field: Deep learning was necessary to achieve 100% in visual generation | Dylan Field | Oct 2, 2025 | — |
| 85% | “85%” | Slack: Competitors discounted products up to 100% to win deals against Amp | Quinn Slack | Sep 25, 2025 | — |
| 85% | “85%” | Huber: Regex handles 90% of code queries; embeddings add marginal improvement | Jeff Huber | Aug 19, 2025 | — |
| 85% | “85%” | Fanelli: Series A Startups Have 80% to 85% AI-Generated Code | Alessio Fanelli | May 7, 2025 | — |
| 85% | “85%” | Beauchamp: Spotify would retain 85% of users with only five artists | William Beauchamp | Jan 26, 2025 | — |
| 85% | “85%” | Reddy: Flagship OpenAI API costs fell 80% to 85% in roughly 18 months | Pranav Reddy | Dec 21, 2024 | Supported |
| 83% | “83%” | Younis: 83% of Applied Intuition's workforce is engineering | Qasar Younis | Apr 27, 2026 | — |
| 80% | “80%” | Beam: Lila's AI hits 80% zero-shot on gene editing, beating humans' 0% | Andy Beam | Jul 16, 2026 | — |
| 80% | “80%” | Malde: In legal AI, getting 80% of the way there is zero | Ronak Malde | Jun 21, 2026 | — |
| 80% | “80%” | Azhnyuk: FPV Drones Cause 70% to 80% of Frontline Casualties | Yaroslav Azhnyuk | May 18, 2026 | Supported |
| 80% | “80%” | Horthy: AI demos work at 80% accuracy, enterprise software needs far more | Dex Horthy | Mar 6, 2026 | — |
| 80% | “80%” | Glaese: OpenAI no longer trusts further score improvements on SWE-bench Verified | Mia Glaese | Feb 23, 2026 | — |
| 80% | “80%” | Wang: Compute accounts for roughly 80% of foundation model fundraises | Sarah Wang | Feb 19, 2026 | — |
| 80% | “80%” | Reggio: Brex targets 80% automated application decisions within 60 seconds | James Reggio | Jan 17, 2026 | — |
| 80% | “80%” | General Intuition aims to drive 80% of AI physical interactions by 2030 | Pim de Witte | Dec 6, 2025 | — |
| 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 | — |
| 80% | “80%” | Taskaya: 80% of promotional video content will be AI-generated within 12 months | Batuhan Taskaya | Sep 8, 2025 | — |
| 80% | “80%” | Dax Reed notes developers avoid context limits via frequent session restarts. | Dax Reed | Aug 5, 2025 | — |
| 80% | “80%” | Brockman: OpenAI's 80% o3 price cut yielded neutral or positive revenue | Greg Brockman | Aug 15, 2025 | — |
| 80% | “80%” | Mohan: Over 80% of software developers are on Windows | Varun Mohan | Jul 28, 2025 | Contradicted |
| 80% | “80%” | Anthropic finds multi-agent architecture outperforms single-agent baseline by 80% | Dylan Davis | Jul 5, 2025 | Partly supported |
| 80% | “80%” | Alberti: 80% to 90% of DeepWiki users view pre-indexed popular repos | Silas Alberti | May 21, 2025 | — |
| 80% | “80%” | Developers will eventually spend 80% of their time controlling agents outside IDEs | Guy Gur-Ari | Apr 2, 2025 | — |
| 80% | “80%” | Agarwal: Synthetic data distillation achieves 80% to 90% of target gains | Rishabh Agarwal | Mar 23, 2025 | — |
| 80% | “80%” | Husain: 80% of LLM-as-a-judge implementations are unhelpful | Hamel Husain | Mar 13, 2025 | — |
| 80% | “80%” | Kilpatrick: Multimodal Live API provides 80% of Project Astra experience | Logan Kilpatrick | Feb 28, 2025 | — |
| 80% | “80%” | Mohan: Over 80% of software developers are on Windows | Varun Mohan | Dec 13, 2024 | Contradicted |
| 80% | “80%” | Howard: 80% of top unique creators have unconventional or non-mainstream backgrounds | Jeremy Howard | Aug 17, 2024 | — |
| 80% | “80%” | Kanjun Qiu: Agent reliability's final 20% is as hard as self-driving cars | Kanjun Qiu | Oct 21, 2023 | — |
| 80% | “80%” | Lambert: GPT-4 achieves 80% preference labeling agreement versus 70% for humans | Nathan Lambert | Jan 11, 2024 | Supported |
| 80% | “80%” | Hotz: Halving GPU power yields 80% of peak performance | George Hotz | Jun 20, 2023 | Supported |
| 77% | “77%” | Swyx: Microsoft and OpenAI account for 77% of CoreWeave revenue | Michael Swix (Swyx) | Apr 11, 2025 | Partly supported |
| 75% | “75%” | Chan: CELLxGENE corpus reached 125M cells with 75% community contribution | Priscilla Chan | Nov 6, 2025 | Supported |
| 75% | “75%” | Landgraf: 75% of internal Ona pull requests are co-authored by AI | Johannes (Johannes Landgraf) | Sep 1, 2025 | — |
| 75% | “75%” | 75% of DeepSeek-R1 usage at a major inference provider is for distillation | Stephanie Palazzolo | Aug 6, 2025 | — |
| 75% | “75%” | Up to 75% of Future UX Interfaces Will Be Voice-Driven | Kwindla Hultman Kramer | May 6, 2025 | — |
| 75% | “75%” | OpenAI increases prompt caching discount from 50% to 75% on GPT-4.1 | Michelle Pokrass | Apr 15, 2025 | Supported |
| 75% | “75%” | Husain: 80% of LLM-as-a-judge implementations are unhelpful | Hamel Husain | Mar 13, 2025 | — |
| 75% | “75%” | Goyal: Over 75% of Braintrust Eval Users Use TypeScript SDK | Ankur Goyal | Oct 11, 2024 | — |
| 75% | “75%” | Lambert: RLHF reward models achieve only 65% to 75% validation agreement | Nathan Lambert | Jan 11, 2024 | Supported |
| 74% | “74%” | Wang: Daytona reports massive 74% month-over-month growth | Shawn Wang | May 21, 2026 | Supported |
| 71% | “71%” | Azhnyuk: Terminal guidance boosted Ukrainian drone pilot's success rate to 71% | Yaroslav Azhnyuk | May 18, 2026 | — |
| 70% | “70%” | Midha: Best-in-class AI cluster MFU is between 60% and 70% | Anjney Midha | Jun 18, 2026 | Partly supported |
| 70% | “70%” | Fredrikson: Skilled red teamers phish human participants 60% to 70% of the time | Matt Fredrikson | Jun 22, 2026 | Open |
| 70% | “70%” | Cooper: Railway achieves roughly 70% margins on bare-metal infrastructure | Jake Cooper | May 20, 2026 | — |
| 70% | “70%” | Stay Sassy EM: 70% of executive work is standard and automatable | Stay Sassy EM | Apr 13, 2026 | — |
| 70% | “70%” | Dean: Early 2B-parameter Google Brain model cut ImageNet-22K error by 70% | Jeff Dean | Feb 12, 2026 | Partly supported |
| 70% | “70%” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 70% | “70%” | Ravi: Over 70% of SAM 3 Dataset Annotations Are Negative Phrases | Nikhila Ravi | Dec 18, 2025 | Supported |
| 70% | “70%” | Sands: 65% to 70% of Stripe engineers use AI coding assistants daily | Emily Glassberg Sands | Oct 30, 2025 | — |
| 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 |
| 70% | “70%” | Sohmers: Positron hardware achieves 70% higher performance than NVIDIA at lower power | Thomas Sohmers | Aug 18, 2025 | Open |
| 70% | “70%” | Swix: Copilot report estimates 60-70% of AI-generated code is checked in | Shawn Wang | Jul 28, 2025 | Contradicted |
| 70% | “70%” | Crivello: 60% to 70% of user agent prompts are meaningless | Florent Crivello | Nov 15, 2024 | — |
| 70% | “70%” | Polu: Dust averages 60% to 70% weekly active penetration in enterprise accounts | Stanislas Polu | Nov 11, 2024 | — |
| 70% | “70%” | Hsu: 60% to 70% of AI business utility comes from ChatGPT | David Hsu | Feb 7, 2024 | — |
| 70% | “70%” | Lambert: GPT-4 achieves 80% preference labeling agreement versus 70% for humans | Nathan Lambert | Jan 11, 2024 | Supported |
| 67% | “67%” | Levie: Box serves 67% of Fortune 500 companies | Aaron Levie | Mar 5, 2026 | Supported |
| 66% | “66%” | Cosine's Genie achieved roughly 66% codebase retrieval accuracy across benchmark tasks | Alistair Pullen | Aug 22, 2024 | Supported |
| 65% | “65%” | Feinberg: Small molecules make up 65% of FDA-approved drugs | Evan Feinberg | Jun 30, 2026 | Supported |
| 64% | “64%” | Shawn Lewis: o1 agent achieves 57% single-pass, 64% with parallel rollouts | Shawn Lewis | Jan 28, 2025 | Supported |
| 60% | “60%” | Park: Frontier models hit only 20-30% accuracy predicting niche human behavior | Joon Sung Park | Aug 21, 2026 | — |
| 60% | “60%” | White: AI hits 60-70% on Bixbench, matching human expert agreement | Andrew White | Jan 28, 2026 | Partly supported |
| 60% | “60%” | Reggio: Brex Card Product Drives 60% of Direct Revenue | James Reggio | Jan 17, 2026 | — |
| 60% | “60%” | Arena sampled open-source models at 60/40, debunking Leaderboard Illusion paper | Anastasios Angelopoulos | Dec 31, 2025 | Supported |
| 60% | “60%” | Corbitt: 55-60% chance RL becomes the standard pattern for deploying scale agents | Kyle Corbitt | Oct 16, 2025 | — |
| 60% | “60%” | Up to 75% of Future UX Interfaces Will Be Voice-Driven | Kwindla Hultman Kramer | May 6, 2025 | — |
| 60% | “60%” | Robicheaux: Florence-2 Achieves 60% mAP on COCO | Peter Robicheaux | Dec 22, 2024 | Supported |
| 60% | “60%” | Reddy: OpenAI's share of enterprise LLM spend dropped from 90% to 60% | Pranav Reddy | Dec 21, 2024 | Supported |
| 60% | “60%” | Crivello: WordPress is a commercial failure despite powering 60% of the web | Florent Crivello | Nov 15, 2024 | — |
| 58% | “58%” | Sands: 58% of Lovable's Payment Volume Flows Through Stripe Link | Emily Glassberg Sands | Oct 30, 2025 | — |
| 57% | “57%” | RFT boosted o3-mini to 57% F1 on medical coding versus clinicians' 40% | Brendan Fortuna | Jul 29, 2025 | Partly supported |
| 57% | “57%” | Shawn Lewis: o1 agent achieves 57% single-pass, 64% with parallel rollouts | Shawn Lewis | Jan 28, 2025 | Supported |
| 56% | “56%” | Baseten's vision-retrofitted GLM-5.2 scored 56% on MMLU Pro without text degradation | Philip Kiely | Aug 3, 2026 | Partly supported |
| 55% | “55%” | Swix: OpenAI Structured Outputs cut API costs by 55% vs Instructor | Shawn Wang | Sep 17, 2024 | — |
| 50% | “50%” | Burazin: RL workloads will reach 50% of Daytona's volume this month | Ivan Burazin | May 21, 2026 | — |
| 50% | “50%” | Chatbase customer model usage is split 50% OpenAI, 50% Anthropic and Google | Yasser Elsaid | May 2, 2026 | — |
| 50% | “50%” | Shah: Supermemory outperformed Claude Code and OpenClaw benchmarks by almost 50% | Dhravya Shah | Mar 9, 2026 | Supported |
| 50% | “50%” | Welling: Perovskite-Silicon Solar Panels Can Theoretically Reach 50% Light Capture | Max Welling | Feb 25, 2026 | Supported |
LATENT SPACELEDGER All shows Try the podcast clipper →