Board

How The Internet Works

Technology180 insights0 followers

@technology
Uber was fundamentally strong while leadership instability, board conflict, controversy, competition, and rapid change created chaos.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Autonomous-vehicle providers can build their own brands while using aggregators to fill capacity and increase fleet utilization.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
You can pursue extreme ambition without letting work and worldly chaos rewrite your identity.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Travel, food delivery, and transportation platforms let suppliers compete for customers while still relying on them for incremental demand.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Protecting children from every difficulty feels kind now, but can leave them weaker later.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Vehicle hardware costs typically fall 30–40% per generation, pushing autonomous transportation toward affordability.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Uber’s recovery became manageable through four initiatives: governance, trust, communication, and the right leadership team.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Uber completed about 1.5 billion trips for people outside their home city last year, with 15% involving airports.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Uber predicts a rider’s destination about three-quarters of the time, reducing the booking interaction to one tap.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
The shock is not that technology feels magical, but how quickly users reset magic into normal.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Uber’s feed and search models are now about 10,000 times larger, enabling predictions built on far more personal data.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Physical AI, including autonomous vehicles and drones, is set to create another trillion-dollar marketplace.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Create randomness in interactions and information flow to bypass organizational filtering.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Memberships work best when serving another member costs little, using fixed assets like empty seats or room upgrades.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Getting users into a multi-service app is not enough. They must discover what it offers.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Drones could cut delivery times from 25–30 minutes to 10–15 minutes while lowering costs and changing food-ordering habits.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
One autonomous vehicle may drive three to four times as much as a human-driven vehicle.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Autonomous vehicles on Uber’s network are at least 30% busier than comparable vehicles outside it.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Developers in India suddenly drove ten times as many code commits using autonomous agents.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Uber burned through its entire annual AI budget in a single quarter.
Invest Like The BestUber CEO on AI, Autonomous Vehicles, and the Future of Transportation
@technology
Private-company diligence can estimate profitability from transaction volume, take rates, employee counts, and comparable-company analysis without full financials.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
The AI stack runs from power and chips through cloud infrastructure and foundational models to applications.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Every Nvidia chip or rack uses 50% to 125% more power, lifting average selling prices for Delta and Advanced Energy.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
The best mentors pair intelligence and wisdom with humility, warmth, grace, and accessibility when people bring difficult problems.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Most of roughly 800 million AI users still use AI 1.0 as a search engine on steroids, not an integrated agentic system.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Replacing copper with fiber for GPU scale-up connections could multiply Corning’s opportunity two to three times.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Supplier relationships are shifting from on-demand purchasing to four-year roadmap collaboration, improving revenue visibility and strategic entrenchment.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Electric vehicles hit an adoption ceiling around 10–15%, far below the earlier expectation that 40–50% of cars would go electric.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
At strategic inflection points, historical data breaks down. Investors need intuition, anecdotes, visual patterns, and connected observations.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Celestica holds roughly 50–60% of the cloud Ethernet switch market, a crucial position as AI workloads become exceptionally network-intensive.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
AI leadership requires scale because only a limited number of companies can afford advanced-model compute.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Rising demand, shortages, prices, and margins can turn low-margin markets into 35–50% annual top-line growth stories for several years.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
On the internet, leadership compounds. The leader becomes bigger, grows faster, and reinforces its position.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
When an expensive AI server fails, the whole system goes down, making essential-component suppliers difficult to replace.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
At roughly 30–40% penetration, exponential growth fades, analysts catch up, and large earnings surprises become less likely.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
AI adoption could continue after foundational models plateau, but commoditization and open-source catch-up could weaken model stocks.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
At $20,000 to $30,000 annually per coder and roughly 20 million coders worldwide, coding implies a potential half-trillion-dollar market.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
AI infrastructure companies can grow units, layer counts, prices, and gross margins simultaneously, multiplying earnings power beyond unit growth alone.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
High-bandwidth memory stacks ten chips, delivers ten times prior input/output performance, and requires years of coordination with GPU designers.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
A single Microsoft data center reportedly contains enough Corning fiber to circle the world four and a half times.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Only 10 basis points of people seriously use AI, yet compute is already sold out and Anthropic reportedly has half what it needs.
Invest Like The BestWhy the AI Boom Is Just Getting Started
@technology
Heavy reliance on coding agents can make engineering work lonely and reduce ordinary team interaction.
Lenny's PodcastWhat AI coding costs engineers
@technology
Engineering teams use shared maker time deliberately to preserve interaction and cohesion.
Lenny's PodcastWhat AI coding costs engineers
@technology
Regular pair-programming lunches help engineers learn one another’s workflows with coding tools.
Lenny's PodcastWhat AI coding costs engineers
@technology
Fast local communication cannot rescue an AI system when communication between chips remains slow.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Pre-fetching turns waiting time into execution by finishing every independent task before the missing component arrives.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Long-tenured engineering teams draw confidence from hard-won experience that can overwhelm newcomers facing the same problems.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
AI hardware requires funding the chip, boards, interconnects, cooling, networking, and the entire cluster.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Building a chip better than every previous AI chip requires enough naivety to challenge what experts accept.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
The strongest hardware teams combine seasoned scale experts with young builders whose chips-on-their-shoulder drive creates extraordinary execution energy.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Pre-fetching cut the time from receiving silicon to running inference from a competitor’s 10 months to 40 days.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Advanced AI will look less like a machine and more like a giant distributed brain spanning countless chips in one data center.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
High-speed AI decoding depends primarily on moving data, not solving harder mathematics, across and within chips.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Repeated contrarian bets get harder to defend emotionally as the stakes grow larger.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Project-based recruiting maps the hardest technical problems, finds who solved them, and tracks those people through repeated conversations.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
The 40-day result came from making everything possible routine before the chip arrived, leaving only final assembly and minor adjustments.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Aggressive spending makes sense when every day without shipping leaves substantial opportunity on the table.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Aggressive fundraising got the rack to market quickly. Existing investors later doubled or tripled down.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
TSMC funded a manufacturing experiment, confirmed its yield improvement, then applied the change across the production line.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
High-frequency-trading engineers were the first believers in kernels-first programming, because they already bypassed compilers.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Inter-chip communication on current NVIDIA products takes roughly 4,000 nanoseconds, against a physical lower bound near two or three.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
A custom interconnect stack above Ethernet’s second layer cuts chip-to-chip latency more than fivefold.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
The first generation of this low-voltage inference technology runs at less than half the voltage of any other AI chip.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Fixing the chip required aligning two clock signals within 50 picoseconds on every chip, two billion times per second.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
A hidden analog clock-domain failure produced incorrect attention results and escaped a massive digital FPGA verification cluster.
Invest Like The BestThe Two Harvard Dropouts Who raised $800M to take on NVIDIA
@technology
Knitting uses two stitches, while programming uses binary zeroes and ones.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
The move from Vim and terminal debugging to IDEs brought integrated debugging, breakpoints, and multithreaded debugging.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Make new mistakes: learning from mistakes is acceptable, while zero-mistake thinking can signal excessive caution and insufficient speed.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI-generated tests remove the upfront tax that made test-driven development impractical for many teams.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Engineers need product sense alongside coding ability to judge whether their work actually improves the product.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI coding has removed the bottleneck and raised the ceiling on what one person can accomplish.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Software delivery escaped rigid CD-manufacturing deadlines when teams gained the ability to ship software online.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Leaders using the product tell teams their work matters and leadership remains engaged, not detached.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
JIT planning keeps roughly one month of priorities visible, then tests them against reality every week.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI can search an unstructured folder for a specific document and perform informal competitive analysis for a small business.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Starting as an individual contributor gives new hires valuable time inside a company’s codebases, tools, languages, and shipping practices.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
A painful, bleeding rash on the nose returned after hotel shampoo and improved after switching to organic shampoo.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI-era teams need creative builders with product sense and deep experts for technically difficult areas.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI is loosening engineering’s grip on the roadmap: product managers can sometimes implement features themselves.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Small-business owners work brutally hard on narrow margins while spending hours on invoicing, expenses, and other disliked administration.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Planning can create alignment today and become obsolete tomorrow when the surrounding environment changes.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Build at the edge of what models almost do. When they catch up, you may lead those who waited.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI shifts engineering ambition from unblocking small features to asking how ambitious the product can become.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
AI productivity is moving past token maxing toward a harder question: what value did we create, and what did it cost?
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
CD manufacturing made engineering time scarce and planning essential. Online delivery and AI-assisted coding moved the bottleneck to verification.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
Anthropic engineers now ship roughly eight times as much code per quarter as they did between 2021 and 2025.
Lenny's PodcastWhat happens after coding is solved? | Fiona Fung (Claude Code & Cowork)
@technology
YouTube initially gave its recommendation algorithm a single goal: increase watch time.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Optimizing YouTube for shares suppresses private topics and loyal niche audiences that rarely share anything publicly.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Creators want YouTube to recommend their videos even when audiences do not want to watch them.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Human moderation after complaints can preserve legal protections. Human approval of everything can make YouTube legally responsible as a publisher.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
A recommendation system can boost watch time immediately while promoting harmful content and creating long-term risks for advertisers.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Neural networks excel when games have obvious scores. Recommendations must first decide what “winning” means for human behavior.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
YouTube cannot inspect or repair an opaque recommendation model. It can only change the feedback signals guiding it.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Watch-time optimization cannot recognize truth, quality, context, or safety. Clicks can therefore reward harmful and conspiratorial videos.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
YouTube’s ideal black box would optimize ad revenue while judging social, financial, reputational, child-safety, and truth-related consequences.
Tom ScottWhy The YouTube Algorithm Will Always Be A Mystery
@technology
Online authority often comes from audience size, not deep research, broad expertise, or superior qualifications.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
When platforms reward videos for keeping viewers on-platform, conspiracy theories and clickbait gain an attention advantage.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
A mostly unedited ten-minute monologue to camera became a popular recurring series despite television conventions demanding movement, cuts, and spectacle.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
YouTube retention graphs reveal opening drops, boredom-driven decline, and an end-card plunge. The algorithm favors gentler declines.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
When a creator shifts from personal videos to science communication, life-focused viewers leave and must be replaced.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Isaac Asimov’s 1950 story about machines quietly manipulating economies to protect humanity makes benevolent optimization deeply unsettling.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
A successful platform cannot be all clickbait or contain none: either extreme drives away users, advertisers, or both.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
YouTube’s advertiser-suitability system linked LGBT-related content with explicit sex, making videos about being gay more likely unsafe.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Apparently, Google serves more advertisements to users who tolerate them, so skipping ads can reduce future ad frequency.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
In a 2015 Reddit survey, users’ biggest reason for not recommending Reddit was exposing friends to hateful content.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
An ideal social algorithm would suppress conspiracy theories and fake news while preserving enough nonsense to keep people engaged.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Giving a video to outside experts can improve its substance while hurting retention, sharing, and algorithmic performance.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Watch-time optimization rewarded longer videos that buried important material at the end, pushing lower-quality videos into recommendations.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Policies protecting vulnerable people from coordinated abuse can also suppress public pushback against dangerous claims and sellers.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Poor or biased training data make machine-learning systems reproduce distortions in their classifications.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
YouTube receives 500 hours of video every minute. Fully human-first moderation cannot operate at that scale.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
AI read retinal photographs to infer sex with 97% accuracy and age within four years.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
A platform can elevate authoritative voices without establishing that those voices are true. Authority and truth are distinct.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Parasocial relationships connect viewers emotionally to performers who do not know them, creating a power imbalance, not reciprocal friendship.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Opaque algorithms work like Skinner boxes: unpredictable rewards keep creators experimenting and breed superstitions about what the system favors.
The Royal InstitutionThere is No Algorithm for Truth - with Tom Scott
@technology
Explaining a model’s prediction means finding the musical features that caused it to choose rock or reggae.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
Would people still hear blues if the same musical pattern played through a radically different instrument?
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
An AI model can appear intelligent while learning a shortcut instead of the information humans expect it to use.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
To find influential frequencies without testing every combination, researchers split them into chunks and repeatedly narrow the search.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
A highly accurate music model can succeed without understanding the musical concept people think it recognizes.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
The Clever Hans effect is a system solving the task by reading an involuntary signal linked to the answer.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
A sufficient signal is any set of frequencies that produces the target classification by itself.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
After removing the sufficient blues signal, the remaining audio was labeled hip hop even though listeners heard something else.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
Blend 100 individually sufficient blues signals and the model still classifies the result as blues, despite muffled ambient noise.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
Change equalization or add delays, and a music classifier can change genre labels while people still recognize the song.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
An audio classifier labeled residual sound hip hop with 83% confidence, while human listeners disagreed.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
A music classifier identified an extremely sparse, unnatural frequency collection as blues, though it sounded nothing like blues to people.
ComputerphileWhy AI is like a (Clever Hans) Horse - Computerphile
@technology
A linked list stores each item with a pointer to the next item instead of relying on indexes.
Fireship100+ Computer Science Concepts Explained
@technology
A recursive function without a base condition continues indefinitely and can cause a stack overflow.
Fireship100+ Computer Science Concepts Explained
@technology
A shell accepts text commands and returns output through a command-line interface to the operating-system kernel.
Fireship100+ Computer Science Concepts Explained
@technology
A variable gives reusable names to data values stored somewhere in computer memory.
Fireship100+ Computer Science Concepts Explained
@technology
Backtracking algorithms test possible options incrementally, returning to alternatives when a path fails.
Fireship100+ Computer Science Concepts Explained
@technology
A while loop repeats until its condition becomes false. A for loop commonly visits every item in an iterable.
Fireship100+ Computer Science Concepts Explained
@technology
Every recursive algorithm needs a base condition that tells it when to stop.
Fireship100+ Computer Science Concepts Explained
@technology
Programming languages make computers practical by hiding lower-level systems behind simpler layers of abstraction.
Fireship100+ Computer Science Concepts Explained
@technology
Character encodings such as ASCII and UTF-8 turn typed characters into binary values.
Fireship100+ Computer Science Concepts Explained
@technology
Recursion begins when a function calls itself. Without a terminating condition, it becomes an infinite loop.
Fireship100+ Computer Science Concepts Explained
@technology
Passing objects by reference lets multiple variables use one heap object without duplicating its memory.
Fireship100+ Computer Science Concepts Explained
@technology
Dynamic programming stores solved subproblems through memoization instead of computing the same results repeatedly.
Fireship100+ Computer Science Concepts Explained
@technology
Data structures organize data. Algorithms are the code that uses it to solve problems.
Fireship100+ Computer Science Concepts Explained
@technology
A function is named code that accepts inputs, performs an operation, and returns an output.
Fireship100+ Computer Science Concepts Explained
@technology
A hash, map, or dictionary finds values through user-defined keys instead of integer indexes.
Fireship100+ Computer Science Concepts Explained
@technology
A stack serves the newest item first. A queue serves the oldest item first.
Fireship100+ Computer Science Concepts Explained
@technology
The company is just over five years old, yet its first revenue dollar arrived only three years ago.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
Rapid growth came from leaps in model intelligence paired with products built around those models.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
A frontier model became available to the business for the first time in March 2024.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
By the end of 2024, the business had raised a Series E and reached nearly $1 billion in run-rate revenue.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
Exponential-growth businesses demand a shift from linear, incremental planning to belief in radically different growth potential.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
The mechanism behind model-led growth is not models alone, but products, go-to-market execution, and distribution.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
Investors saw $1 billion of run-rate revenue as achievable, then doubted another tenfold given enterprise adoption timelines.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
The company jumped from roughly $9 billion in run-rate revenue at the year's start to more than $30 billion by quarter-end.
Invest Like The BestIs Anthropic the Fastest Growing Company Ever?
@technology
Every software product encodes human decisions about behavior, appearance, controls, and architecture, then executes its creators’ intentions.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Large code migrations and version upgrades fit agents early: repeated changes across huge codebases still require thoughtful trade-offs.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Most of the gap between today’s coding tools and precise AI control could disappear within roughly five years.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Ranking engineers by tokens consumed confuses activity with productivity. Measure useful work produced instead.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Many white-collar tasks remain manual even when software could handle them. One-off tools are not economically worthwhile.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Early enterprise AI deployment starts hands-on: perform the work manually, understand the workflow, then build the orchestration around it.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Difficult engineering problems justify the strongest reasoning models. Repetitive or boilerplate work favors faster models.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Technology can feel jarring, become normal quickly, and leave people unable to imagine life without it.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Early AI agents dazzled in demos and toy projects, then failed inside real companies and production codebases.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Today’s software-building interfaces are not the endpoint. Many generations of new product experiences remain to be created.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
A hatred of losing can begin in early childhood and remain a stable personality trait across every competitive arena.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
The founding team sees this company as its big one, aiming to build a generational business and hyperscaler despite uncertain success.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Devin’s first task was setting up MongoDB and diagnosing the successive errors that appeared during the initial database setup.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Teams using Devin reportedly ship ten times faster and do ten times more.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Deep domain focus can beat larger generalists when the focused company cares more about a problem’s practical nuances.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Losing after giving everything hurts. Discovering your team could have pushed further and chose not to hurts more.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
A scoped AI initiative could replace an 18-month, $15 million outsourced project with a $1 million internal effort in three months.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
AI-generated software could make one-time and highly specialized automation economically viable at last.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
A year-long AI agent should pursue consequential missions, not waste its defining advantage formatting bulk email.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
An AI’s working horizon changes delegation itself: seconds mean commands, hours mean tasks, and years mean missions.
David SenraBuilding The World's First AI Software Engineer | Cognition’s Scott Wu
@technology
Telegraph networks grew from experiments connecting two ends of a garden to neighborhoods, train signaling, and multiple cities.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
Undersea cables protect their fragile fibers from seawater with only a few layers, including petroleum jelly.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
Today’s global digital network still rests on technology dating back 157 years: submarine telegraph cables.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
Some undersea cables descend as deep as Mount Everest is high, far below ordinary sea life.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
Installation ships plow grooves into the ocean floor, lay cables inside, and let currents bury them with sand.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
About 300 undersea fiber-optic cables carry 99% of international data traffic, making oceans the internet’s hidden backbone.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
The internet’s ocean infrastructure is vulnerable to earthquakes, ship activity, and even confused sharks biting cables.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
Satellites, balloons, and cell towers reach above sea level, yet Facebook and Google still choose undersea cables for speed.
VoxThin underwater cables hold the internet. See a map of them all.
@technology
The first submarine cables used exotic tree sap as waterproofing before petroleum jelly became standard.
VoxThin underwater cables hold the internet. See a map of them all.

Keep what matters

Save insights to your own Boards and return to them anytime.