The Algorithmic Bridge: Year One

Stats: How TAB has grown over time / Survey: What do you think about TAB / Coming: More optimism! More themed series! / Archive: All TAB posts, categorized

The Algorithmic Bridge has turned one this week.

I can now confidently say what I already suspected from the very beginning: Moving to Substack and creating TAB has been the best decision I've made in my short 3-year career as a writer.

Previous to this I was trapped in an algorithmically-powered game that decided, rather obscurely, whether my beloved readers would find my writing. That was Medium. My home for a year and a half.

It was a tough decision to leave (I only republish free articles over there now) because I didn't know what was awaiting me outside, but it turned out GREAT. Substack has been an amazing virtual home for me. And you guys have been the best housemates I could've asked for. As long as this place is still standing, I will live here.

I want to thank all of you who made time to comment on last week's open thread (it's still open for anyone who wants to introduce themselves or add to the conversation!) It's nice getting to know you little by little.

That said, I want to share a few things with you today about my one-year journey on Substack (and a few other things about the near-term future!)

Stats: How TAB has grown over time

I love numbers and monotonically-growing charts and I suppose you love them, too, so I share below the main ones that reflect the health of the newsletter: free and paid subs (with key milestones), ARR, all-time views, and TAB around the world.

I know a significant fraction of you also write on Substack, so I hope my progression may help you figure out what to expect if you're getting started (if you’re in this situation, feel free to reach out with any questions you may have).

I brought with me 700 emails I gathered in one and a half years on Medium. One year later, we’re now at almost 13,000! I’m not there yet—at a point where I can lay back and feel I’ve achieved my goals—but the steady progress and projection I see encourage me to continue as strong as the very first day.

Key milestones

Virality spikes matter for long-term growth

California: The epicenter of AI—and TAB’s success. Also, hey North Dakota!

TAB is global: 3 out of 4 countries read it. Also, 1/3 of you are Americans!

Survey: What do you think about TAB

If there’s one way that I can further improve TAB so that it better fits your expectations and provides the value you’re looking for, is by asking you directly.

Taking inspiration from Noah Smith, one of Substack's most-read economy bloggers, I created a feedback anonymous survey for subscribers.

It’ll help me gather your opinion in a more formal/standardized way so that I can steer TAB toward our collective arithmetic mean (nerd!). It takes ~2 min (9 questions).

Coming: More optimism! More themed series!

Here are two aspects I want to focus on in the future, to transform TAB into a place I—and hopefully you—feel more comfortable with.

• More optimism! This first year I’ve focused TAB on fighting AI overhype. I am aware that the topics I’ve chosen—and especially the approaches I’ve taken—can feel somewhat negative at times. Now that I’ve established my position as anti-hype, I feel I’m free to delve unapologetically into my long-jailed optimism. TAB will be a brighter place from now on (I won’t neglect my critical self, though). I’m confident you want more of that even before checking the results of the survey ;)

• More themed series! The benefit is twofold:

  • First, a series—in contrast to standalone posts—provides richer coverage and a more robust structure to broad topics that otherwise lack cohesion and coherence. If anything, this will benefit my intention to provide a bird’s-eye view of AI without making the content too technical or the articles intolerably lengthy.

  • Second, it takes away the burden of coming up with sufficiently varied topics every week so that neither you nor I get bored (overarching themes have enough heterogeneity within for this to work just fine). I’ll interleave non-related articles, too—gotta find the balance between timely and evergreen content!

Archive: All TAB posts, categorized

Finally, and again taking inspiration from another Substack heavyweight, Gergely Orosz (the absolute leader of the technology category) I classified below all the posts I've published on TAB in its first year (June 2022 - June 2023) so that you can check them at a glance.

They’re ordered by category and then chronologically (higher in the list = more recent). In total, I’ve published 139 articles (those with a 🔒 are for paid subs), which amounts to 2,5 articles per week on average. The word count is between 250-300,000 words, so around 3-4 books’ worth of reading!

ChatGPT, GPT-4, and Language Models

• Size Doesn't Matter 🔒

• ChatGPT: Eraser of the Implausible

• GPT-4 in 10 Keys 🔒

• 5 Practical Applications Where ChatGPT Shines

• Why I Don't Buy the 'ChatGPT = Calculator for Writing' Analogy 🔒

• GPT-4: A Viral Case of AI Misinformation

• ChatGPT and the Future (Present) We're Facing 🔒

• How to Get the Most Out of ChatGPT

• OpenAI Has the Key To Identify ChatGPT's Writing 🔒

• ChatGPT Is the World’s Best Chatbot

• What Does an AI Say to Another?

• Galactica: What Dangerous AI Looks Like

• GPT-4 Rumors From Silicon Valley 🔒

• Ultra-Large AI Models Are Over 🔒

• OpenAI Whisper Holds the Key to GPT-4

• Google Is Making LaMDA Available

• GPT-3 Wrote a Paper About Itself

• BLOOM Is the Most Important AI Model of the Decade

Business Wars (Google, Microsoft, OpenAI, DeepMind)

• OpenAI Rushed Up the Ladder It Is Now Kicking Away 🔒

• How OpenAI Fooled Us All 🔒

• Why Bing Chat Is the Most Important Failure in AI History 🔒

• Google vs Microsoft (Part 3): A New Way of Doing—and Experiencing—AI

• Google vs Microsoft (Part 2): Google's Bard May Prove Everyone Wrong

• Google vs Microsoft (Part 1): Microsoft’s New Bing Is a Paradigm Change for Search and the Browser

• DeepMind's Sparrow vs OpenAI's ChatGPT 🔒

• Microsoft vs Google: Will Language Models Overtake Search Engines?

• An OpenAI Story: From Hegemon to Underdog

Social and Cultural Effects of Generative AI

• Why No One Wants AI Ethics Anymore

• How to Stop AI FOMO 🔒

• AI Can’t Benefit All of Humanity

• The World Must Reconcile AI Skepticism and AI Optimism 🔒

• Who's to Blame for AI-Generated Harm—Users or Companies?

• AI Influencers From the Post-ChatGPT Era

• How AI Can Help With the Loneliness Epidemic

• Generative AI: My Enhancement, Your Replacement 🔒

• The AI Empathy Crisis

• Do You Also Feel AI Is Going Too Fast? 🔒

• AI Is a Celebrity Technology 🔒

• When AI Can Revive the Dead

• GPT-4Chan: 'The worst AI ever'

• AI Has an Invisible Misinformation Problem

Philosophical Reflections

• AI’s Greatest Lie 🔒

• (M)ending the World

• How to Remain Human in the Age of AI 🔒

• Schrödinger's AGI 🔒

• GPT-4: The Bitterer Lesson

• On the Dangers of Overused AI Metaphors

• The Alienness of AI Is a Bigger Problem Than Its Imperfection 🔒

• The Four Filters of AI Knowledge

• AI’s Destiny Was Never Intelligence 🔒

• Why AI Is Doomed Without Neuroscience

• Life After Work

• How AI Could Be the End of the Internet as We Know It 🔒

• A World of Homo Sapiens and Machina Sapiens 🔒

• Why 'Is LaMDA Sentient?' Is an Empty Question

Predictions and AI Risk

• A Non-Cynical Reading of AI Risk Letters

• Superintelligence: OpenAI Says We Have 10 Years to Prepare

• The Missing Piece in Geoffrey Hinton's Newfound Fear of AI Existential Risk 🔒

• Why the FLI Open Letter Won't Work

• 4 Optimistic Takes on the Future of AI 🔒

• Top 7 AI Predictions for 2023 🔒

• Generative AI Could Pollute the Internet to Death 🔒

• Why Elon Musk’s Fear of AI Is Misguided

• From Siri to Photoshop to Google Search—Large AI Models Will Redefine How We Live 🔒

AI and Writing

• Why I (Don't Use ChatGPT to) Write

• A Bull, a Rose, a Tempest

• No Writer Is Safe From AI 🔒

• Why Generative AI Angers Artists but Not Writers

• Google's Wordcraft: An AI Writing Tool Powered by LaMDA

• I Co-Wrote This With an AI Writing Tool 🔒

• Everyone Is Wrong About AI Writing Tools 🔒

• Why I'll Always Disclose My Use of AI Writing Tools

AI Art

• Stable Diffusion 2 Is Not What Users Expected—Or Wanted

• While AI Artists Thrive, Detractors Worry—But Both Miss What's Coming 🔒

• Stable Diffusion Is the Most Important AI Art Model Ever

• How Today's AI Art Debate Will Shape the Creative Landscape of the 21st Century 🔒

• AI Reimagines 10 Famous Landscape Paintings

• OpenAI Is Opening DALL·E 2

• DALL·E Mini Won—But Now OpenAI Is About to Open DALL·E 2

• DALL·E 2 Will Disrupt Art Deeper Than Photography Did 🔒

Demystifying Generative AI Series

• #4 ChatGPT Will *Always* Make Stuff Up—What Can We Do About It? 🔒

• #3 Generative AI Is a Net Loss for Workers 🔒

• #2 Open Source AI Is Not Winning—Incumbents Are

• #1 Generative AI Is Not a Magic Wand 🔒

Prompt Engineering

• 3 Prompt Engineering Lessons You Must Know 🔒

• Learn To Master Prompt Engineering With This Singular (Triple) Framework 🔒

• Prompt Engineering Is Probably More Important Than You Think

Beyond Generative AI

• Tesla AI Day: Optimus Bot Was Better Than Anyone Expected 🔒

• Google’s PaLM-SayCan: The First of the Next Generation of Robots 🔒

• ACT-1: How Adept Is Building the Future of AI with Action Transformers

• AI Emotion Recognition Is a Pseudoscientific Multi-Billion Dollar Industry

• How the Future of Music Creation Could Belong to AI

Social Media

• Twitter, Elon Musk, and the Information Crisis

• Algorithms Know You. Soon, They Could Hack You 🔒

• TikTok Is the Perfect Social Media — Here's Why You Should Delete It 🔒

• In Search of an Algorithm for Well-Being

Open threads

• Ask Me Anything (AMA): 10,000 Subscribers!

• Thread #8: Ask Me Anything (AMA)

• Thread #7: The Algorithmic Bridge on 2023

• Thread #6: Ask Me Anything (AMA)

• Thread #5: Ask Me Anything (AMA)

• Thread #4: AI Art Predictions and Suggestions

• Thread #3: AI & Life After Work

• Thread #2: Large Language Models as Virtual Assistants?

• Thread #1: Ask Me Anything (AMA)

Meta

• One Year of The Algorithmic Bridge!

• The Algorithmic Bridge in 2022

• Welcome to The Algorithmic Bridge

WYMHM: Every major AI event since September, in one place

• What You May Have Missed #31: 🔒 Sam Altman's World Tour Highlights / Research (Orca, AlphaDev, MusicGen) / Products (Google, The Frost) / Articles (AI or nuclear tech?) / Misc (Images as QR, Generative Fill's fixing pics)

• What You May Have Missed #30: 🔒 AI risk statement / AI journalists doing a good job? / Research (Social Turing Test) / Product (courses by Google & Andrew Ng) / Articles (The Illusion of China’s AI Prowess) / Misc (Is GPT-4 worse?)

• What You May Have Missed #29: 🔒 Research (5% US adults find ChatGPT very useful, Falcon, Voyager, MMS, CoDi, QLoRA) / Products (Copilot, Firefly) / Essays (Best: Altman vs EU) / Talks (State of GPT) / Misc (The ChatGPT lawyer case)

• What You May Have Missed #28: 🔒 R&D and products (ToT, DragGAN, LIMA, ChatGPT iOS) / Essays (Best: prof falsely accused students of using ChatGPT) / Miscellanea (fake ID hacks, LeCun's snarky takes) / Senate hearing on AI oversight

• What You May Have Missed #27: 🔒 Google I/O (PaLM 2, Bard, Search, Gemini) / Creatives in check (Brian Merchant, Grimes, AI movies) / Essays (Best: Dylan Patel) / R&D (100K context LMs, LMs to explain LMs) / Misc (Gary Marcus TED)

• What You May Have Missed #26: 🔒 Open source (Google moat, blooming ecosystem) / AI regulation (writers' strike, White House meeting) / Geoffrey Hinton / TED talks / Worthwhile essays / R&D (emergence, MJ v5.1) / Misc and curiosities

• What You May Have Missed #25: 🔒 LLMs (evolutionary tree, LLaMA) / Generative agents (BabyAGI, Auto-GPT) / AI art / Worthwhile essays / R&D and products (Elon Musk's X.AI) / Miscellanea and curiosities (Drake & The Weeknd viral song)

• What You May Have Missed #24: 🔒 AGI / GPT-4 agents / LLMs / Essays against the open letter / LMs hallucinations / Other LM problems / AI regulation / R&D and products / Worthwhile magazine essays / Miscellanea and curiosities

• What You May Have Missed #23: 🔒 Twitter algo / AI regulation / Open letter aftermath / E. Yudkowsky / Deepfakes / Bard / GPT-4 / R&D and products / Worthwhile magazine essays / Miscellanea and curiosities

• What You May Have Missed #22: 🔒 Beyond GPT-4: An eventful week / GPT-4 (security challenges, hallucinations, and misinfo) / Open-source AI is dying / An array of generalized bad sensations about AI (anxiety, exhaustion, FOMO, etc.)

• What You May Have Missed #21: 🔒 GPT-4 coming this week? / Noam Chomsky: The False Promise of ChatGPT / Midjourney, DALL-E, and Stable Diff / Social media + generative AI / Open-source AI / Human brains = ChatGPT? / PaLM-E / Misc

• What You May Have Missed #20: 🔒 ChatGPT API / Generative AI beyond ChatGPT: Meta's LLaMA, Microsoft's Kosmos-1 and Bing updates, OpenAI rival, and Stable Diffusion / Who´s responsible for AI harm, users or companies?

• What You May Have Missed #19: 🔒 The less utopian side of AI / Key developments: LLaMA, Robot ChatGPT, ControlNet & other generative AI news

• What You May Have Missed #18: 🔒 Short overview of Microsoft vs Google and Bing/Sidney / Robust AI and robust testing methods / ChatGPT and the limitations and applicability of LLMs / Other generative AI news

• What You May Have Missed #17: 🔒 Google-Anthropic vs Microsoft-OpenAI / Attacks on OpenAI / 5 generative AI issues that don’t directly involve OpenAI

• What You May Have Missed #16: 🔒 AI writing in journals and magazines / The clash between Big Tech and AI startups / News beyond text generation: Video / Law, security, and porn

• What You May Have Missed #15: 🔒 ChatGPT from the sociocultural perspective / ChatGPT from the business perspective / Miscellanea (because, believe it or not, there’s life beyond ChatGPT)

• What You May Have Missed #14: ChatGPT: Microsoft deal, monetization, fake scientist, education / GPT-4 info and misinfo / More generative AI news: a robot lawyer, Stable Diffusion lawsuit, AI journalist, RLHF for images, 10x Alexa

• What You May Have Missed #13: 🔒 ChatGPT and a whole world of implications / On the future of AI / Bad uses of language models / Meanwhile on the visual generative space

• What You May Have Missed #12: 🔒 On AI art, trad artists, and the clash between the two

• What You May Have Missed #11: 🔒 NeurIPS 2022 / OpenAI and ChatGPT / DeepMind's crazy week / Weekly dose of generative AI / Other AI news: McKinsey, Google, Karen Hao, and Noah Smith+roon

• What You May Have Missed #10: 🔒 Stable Diffusion 2 / Generative AI news / Meta's Cicero and Neurosymbolic AI / AI misinformation / Four bits on automation / Using robots to talk to animals

• What You May Have Missed #9: 🔒 Stability.ai AMA / Stable Diffusion 2 vs Midjourney v4 / More generative AI news / LLMs: From GPT-3 to Galactica / NVIDIA + Microsoft AI supercomputer / Seeing beyond the brain

• What You May Have Missed #8: 🔒 A week of generative AI: a new stage, with critical obstacles, and new ideas (text, audio, and visual)

• What You May Have Missed #7: 🔒 DALL·E API & Midjourney v4 / 3 new Gen AI apps /AI ethicists' hardships / When generative AI becomes unethical / WebSummit: Gary Marcus and Noam Chomsky / Meta protein folding model / Google AI@ '22

• What You May Have Missed #6: 🔒 Self-driving cars aren't coming anytime soon / Is scale all we need? / Personalized avatars with Stable Diffusion + Dreambooth / Image to music / OpenAI partners with Shutterstock

• What You May Have Missed #5: 🔒 Generative AI for music / New unicorns: Stability.ai and Jasper.ai / Gen AI landscape map / Drama in the OS community / New LLM state-of-the-art / Microsoft: GitHub Copilot copyright infringement

• What You May Have Missed #4: 🔒 EU and US AI accountability / DeepMind's AlphaTensor improves algorithms / Bot translates videos into English / Self-Ask to improve GPT-3's Q&A perf / Boston Dynamics rejects robot weaponization

• What You May Have Missed #3: 🔒 OpenAI removes DALL·E waitlist / Bruce Willis' digital twin / Text-to-video & text-to-3D / Stock AI image libraries are now a thing

• What You May Have Missed #2: 🔒 AI could cause a global catastrophe / Kids using AI to write essays / Clearview AI to redeem itself / AI artist copyrighted Midjourney's work / AI pioneers disagree on its future: Is scale enough?

• What You May Have Missed #1: 🔒 Stable Diffusion's monster / EU AI Act / Meta AI wants to read your brain / Fast.ai's new AI course / Anthropic advances AI interpretability / Stable Diffusion's endless possibilities

---

If you’re a writer I hope this overview encouraged you to continue your journey and keep growing. Substack is the best platform for writers right now on the internet.

If you’re a reader, I hope it gave you a glimpse of what TAB looks like from my end. You are the reason TAB exists at all. I would deeply appreciate it if you could, to the extent of your ability, help us grow together: by sharing, recommending, or subscribing.

Let’s go for the second year!