The expertise trade has all the time had a barely dysfunctional relationship with hype. Founders insist they’re impartial thinkers, enterprise capitalists declare they’ll see round corners, after which everybody quietly updates their pitch decks when the most recent narrative begins attracting cash.
A chart produced by Sequoia Capital companion Konstantine Buhler, working with Sequoia intern Oliver Cho, gives an unusually revealing view of this habits. The pair ranked the 15 most mentioned matters on Hacker Information in annually from 2007 to 2026, then positioned probably the most beneficial firm based throughout that yr beneath the corresponding rankings.
Sequoia’s 20-Yr Hacker Information Chart, Supply: X
The outcome resembles a plate of multicolored spaghetti, however the underlying message is remarkably coherent. The loudest topic in any given yr not often produces that yr’s most respected firm. On the identical time, necessary technological actions typically start showing in specialist communities 5 – 6 years earlier than they turn into mainstream. Hype, in different phrases, is often a horrible marketing strategy and a surprisingly helpful radar system.
What the Sequoia chart really measures
The chart tracks the relative recognition of topics mentioned on Hacker Information, rating every subject from primary to quantity 15 for yearly since 2007. Its highlighted classes embrace internet improvement, cloud infrastructure, programming languages, cryptocurrency, social media, Apple, synthetic intelligence, geopolitics and what the authors name the “Musk-verse.”
Beneath the rankings is a second timeline exhibiting what Sequoia considers probably the most beneficial firm based in annually.
That distinction is important. The chart isn’t measuring technological progress, income development, buyer adoption or the quantity of capital invested in every trade. It measures consideration amongst a extremely particular on-line neighborhood.
Regardless of the chart’s title, Hacker Information isn’t “the web.” Its official neighborhood tips describe the location as a discussion board for materials that intellectually curious hackers would discover attention-grabbing. Routine movie star, sports activities, crime and political tales are typically discouraged until they reveal some wider technological or social phenomenon.
That makes Hacker Information a poor illustration of mass tradition however a doubtlessly glorious early-warning system for developer habits. Infrastructure, open-source software program and obscure analysis developments might floor there years earlier than atypical shoppers encounter them. Client merchandise, vogue and leisure traits might seem comparatively late—or by no means.
There are additionally methodological questions that the general public put up doesn’t reply. It doesn’t disclose whether or not “hype” was calculated from article frequency, feedback, factors, engagement or some weighted mixture. Nor does it clarify exactly how discussions have been clustered into broad classes.
The classes themselves aren’t completely comparable. “Programming languages” describes a technical subject. “Apple” is one company. “Geopolitics and battle” covers a sprawling assortment of world occasions. The “Musk-verse” combines a number of corporations, applied sciences, political disputes and one unusually attention-efficient human being.
The chart ought to due to this fact be handled as an insightful visualization, not an econometric proof. It’s glorious mental ammunition and a reasonably blunt scientific instrument.
The businesses that bought constructed
The underside row is sort of a parallel historical past of the fashionable startup economic system. From 2007 by means of 2025, the chart identifies MongoDB, Airbnb, Uber, Stripe, Zoom, ByteDance, Databricks, Tether, OpenAI, Cerebras, Anduril, Kalshi, Ramp, Wiz, Anthropic, Cursor maker Anysphere, xAI, Secure Superintelligence and Considering Machines Lab as probably the most beneficial corporations based of their respective years. The 2026 place stays clean as a result of the winner can not severely be recognized whereas the yr continues to be underway.
The figures vary from roughly $22 billion for Kalshi to $965 billion for Anthropic. A number of of the biggest numbers have recognizable sources: OpenAI announced a $122 billion financing at an $852 billion post-money valuation in March 2026, whereas Anthropic announced a $65 billion round valuing it at $965 billion in Could. Ramp’s $44 billion determine adopted its June 2026 Series F financing, whereas the chart’s $60 billion worth for Anysphere displays SpaceX’s settlement to accumulate the corporate behind Cursor.
These values aren’t apples-to-apples. The row mixes public market capitalizations, non-public post-money valuations, acquisition costs and estimates. A publicly traded firm’s market worth adjustments constantly. A personal valuation could also be established by a comparatively small financing spherical containing investor protections. An acquisition worth might embrace a strategic premium unavailable to atypical shareholders.
The row additionally comprises huge hindsight bias. It’s comparatively straightforward to establish the main firm based in 2008 after observing 18 years of subsequent efficiency. Choosing the definitive winner from 2024 or 2025 is nearer to declaring the Method One champion through the formation lap.
Buhler acknowledges the issue, noting that the highest corporations from the newest years stay undecided. That warning must be printed in daring beneath the whole right-hand facet of the chart.
https://x.com/Konstantine/standing/2085053452640817334
The trade talked about platforms whereas founders constructed functions
The chart’s clearest examples come from the early years. Google was the highest Hacker Information subject in 2008, however that yr’s eventual winner was Airbnb. In 2009, Hacker Information dialogue was dominated by low-level programming and programs engineering whereas Uber was being created.
At first look, this seems to vindicate Buhler’s conclusion that “chasing hype not often results in enduring outcomes.”
The extra attention-grabbing interpretation, nevertheless, isn’t that the technical neighborhood was unsuitable. It’s that the financial worth appeared one layer away from the dialog.
Airbnb was not constructing one other search engine, however it relied on search promoting, digital id, on-line funds and mapping. Uber was not promoting operating-system expertise, however it couldn’t have existed with out smartphones, GPS, cellular broadband, cloud computing and software program infrastructure.
Stripe, based in 2010, didn’t have to invent the net. It wanted to take away one of many internet’s nastiest industrial bottlenecks: accepting funds with out forcing builders to spend weeks wrestling with legacy banking programs.
The profitable corporations have been ceaselessly not promoting the applied sciences being mentioned. They have been exploiting the brand new behaviors these applied sciences had made doable.
That’s the crowd accurately figuring out the muse whereas failing to foretell which constructing can be constructed on high of it.
New traits actually do announce themselves years early
Buhler’s strongest commentary is that “new traits announce themselves 5 to 6 years early.”
Massive language fashions first entered the chart’s high 15 in 2016 however didn’t attain primary till 2022. AI coding and brokers entered at quantity 12 in 2021, fell to quantity 13 in 2022, rose to sixth in 2023 and reached second place in 2024 and 2025. Within the year-to-date 2026 rating, AI coding and brokers lastly overtakes normal LLMs and chatbots.
To the general public, the arrival of ChatGPT seemed nearly instantaneous. To researchers and software program engineers, it was the industrial end result of years of mannequin improvement, scaling, higher {hardware}, new architectures and more and more succesful interfaces.
The identical course of is now taking part in out in AI coding. The primary section of generative AI was primarily conversational: customers requested questions and acquired generated solutions. The rising agentic section is operational. Software program is being requested to examine repositories, write and take a look at code, function instruments, browse programs and full multi-step workflows.
The applying layer is already turning into brutally aggressive. A current Gilded Age analysis of open-source coding models argues that specialised fashions could possibly slender elements of the hole with a lot bigger proprietary programs, doubtlessly shifting sturdy worth towards coaching knowledge, infrastructure, distribution and workflow lock-in slightly than whichever coding interface occurs to be trendy this quarter. The article rigorously notes that some efficiency claims nonetheless require impartial replication, which is exactly the kind of caveat lacking from most AI victory laps.
That is the place founders repeatedly make the unsuitable transfer. They see a quickly rising class and reproduce its most seen product slightly than investigating the enabling infrastructure, unsolved constraints and second-order penalties beneath it. By the point a pattern reaches primary, the apparent thought is often not the great thought.
Crypto was early, explosive and spectacularly unstable
Crypto gives the chart’s purest instance of hype arriving effectively earlier than consensus. Based on Buhler, cryptocurrency first entered Hacker Information’ high 15 in 2011, years earlier than the 2017 preliminary coin providing growth and the market’s second consideration peak in 2021. By 2026, nevertheless, Crypto and Web3 has fallen to quantity 14—above solely house within the newest rankings.
It could be tempting to conclude that the class was merely a speculative obsession that ultimately ran out of consumers. There may be loads of proof for that interpretation, notably throughout the NFT, token and decentralized-everything frenzy.
However falling consideration doesn’t essentially imply falling utility. Applied sciences typically turn into much less culturally attention-grabbing after they start turning into infrastructure.
Stablecoins and programmable funds are more and more being integrated into machine-to-machine commerce and autonomous software program. Brave New Coin has examined the rise of AI agents capable of controlling cryptocurrency wallets, together with the intense prompt-injection, credential and private-key dangers created when autonomous programs are permitted to maneuver cash.
Courageous New Coin has additionally coated the broader argument that AI agents may need crypto and stablecoins to participate efficiently in financial markets, notably when software program should switch worth globally, constantly and and not using a typical checking account hooked up to a human operator.
Crypto might due to this fact be present process the least glamorous type of success: turning into plumbing. The profile-picture carnival, movie star tokens and decentralized dog-food subscriptions attracted the eye. Secure settlement, programmable possession and machine funds might produce the enduring worth.
Persistent “boring” consideration could also be the perfect sign
Net improvement is arguably the chart’s most spectacular class. It stays close to the highest throughout nearly the whole 20-year interval and sits at quantity three in 2026.
Cloud and infrastructure follows the same sample. It by no means receives the identical explosive surge as crypto or generative AI, however it stays persistently necessary and finishes fifth within the newest rating.
That is what real technological absorption seems like. As soon as a expertise turns into foundational, individuals cease treating it as a revolution and start treating it as a requirement.
Cloud computing didn’t turn into much less necessary when founders stopped inserting “cloud-based” into each sentence. It turned the default location the place software program lived. Net improvement didn’t disappear when crypto and AI turned trendy. Each blockchain software, chatbot, coding agent and enterprise mannequin nonetheless requires interfaces, authentication, databases, observability, deployment programs and APIs.
Boring expertise is usually the place the sturdy cash hides.
The chart’s “Musk-verse” gives a unique type of persistent consideration. It has remained within the high 15 for 14 consecutive years, shifting with developments at Tesla, SpaceX, X, Neuralink, xAI and Musk’s increasing political presence.
Buhler argues that sustained on-line consideration is uncommon and due to this fact “tends to mark one thing actual.” That’s broadly appropriate—however relevance and investability aren’t the identical factor.
Persistent dialogue might point out a transformative expertise. It might additionally point out controversy, movie star, political energy or a person with an unmatched skill to transform tweets into information cycles. The Musk-verse is definitely consequential. That doesn’t imply each startup attaching “Musk-adjacent” language to its deck has found a market.
What Hacker Information is obsessive about in 2026
The best-hand facet of the chart affords a compact snapshot of the expertise trade’s present way of thinking.
AI coding and brokers sits at primary, adopted by LLMs and chatbots. Net improvement stays third. Geopolitics and battle ranks fourth, with cloud and infrastructure fifth and US politics sixth.
Low-level programs and programming languages occupy seventh and eighth. Apple is ninth, finance and markets tenth, and authorities and regulation eleventh. The Musk-verse is twelfth, social media thirteenth, Crypto and Web3 fourteenth, and house fifteenth.
The rating displays a expertise trade more and more preoccupied with autonomous software program, mannequin competitors, infrastructure constraints, political intervention and nationwide energy.
It additionally highlights an necessary change within the relationship between hype and firm formation. The current corporations within the backside row are much more intently aligned with the dominant dialogue matters than the early corporations have been.
OpenAI, Anthropic, Cursor, xAI, Secure Superintelligence and Considering Machines are all immediately related to the AI wave. That complicates the chart’s headline conclusion. Within the present cycle, chasing the dominant pattern has produced a number of of the world’s most respected younger corporations.
There are two believable explanations. The optimistic interpretation is that buyers and founders have turn into higher at figuring out structural technological shifts. The much less flattering interpretation is that massive volumes of capital, computing energy and expertise are being concentrated into one dominant narrative, pushing the non-public valuations of just about the whole lot related to AI into the stratosphere.
The chart’s hidden biases matter
The visualization comprises a number of biases that founders and buyers ought to take note.
First, it makes use of survivorship. We see Airbnb and Uber, not the 1000’s of failed startups based alongside them.
Second, it makes use of hindsight. A historic winner is chosen after its trajectory turns into seen, whereas current winners are based mostly on provisional valuations.
Third, it treats consideration as a single variable. Optimistic enthusiasm, technical criticism, outrage, skepticism and mock can all produce excessive engagement. A topic might rank extremely as a result of Hacker Information customers consider it’s going to change the world—or as a result of they consider it’s full nonsense.
Fourth, it measures correlation slightly than causation. An organization being based throughout a yr by which an unrelated topic was in style doesn’t set up that avoiding the favored topic brought on the corporate to succeed.
Lastly, the chart doesn’t seize pivots. The corporate that in the end turns into beneficial might bear little resemblance to the thought its founders initially pursued. Startup historical past is stuffed with founders who entered by means of one market, discovered one thing uncomfortable and constructed a completely totally different enterprise.
This doesn’t make the chart ineffective. It makes it extra helpful when interpreted correctly.
Ought to founders chase hype or ignore it?
Neither. Founders ought to use hype as radar, not as navigation. A subject coming into the underside of the rating and refusing to vanish is usually extra attention-grabbing than a subject abruptly reaching primary. Early, persistent dialogue means that technically succesful individuals are repeatedly encountering a risk or downside they can’t dismiss.
A pointy vertical spike means the remainder of the market has in all probability found it too. Capital turns into simpler to lift, however expertise turns into costlier, opponents multiply and differentiation collapses. Everybody arrives on the gold rush promoting the identical AI pickaxe.
LLMs at quantity 12 in 2016 have been an rising sign. LLMs at primary in 2023 have been consensus. Crypto in 2011 represented an odd new technical motion. Crypto through the 2017 ICO growth was a touring circus with Telegram teams.
The most effective founders aren’t merely asking what’s scorching. They’re asking what has just lately turn into doable, which bottleneck stays unsolved, what infrastructure each new entrant would require and which buyer behaviors will change when the expertise turns into dependable.
The gang is usually unsuitable in regards to the firm that may win. It’s much less ceaselessly unsuitable that one thing necessary has began occurring.
That’s the actual message inside Sequoia’s chart. Hype isn’t reality, and recognition isn’t product-market match. However consideration is knowledge.
Ignoring that knowledge as a result of everybody else is being obnoxious about it’s not contrarian considering. It’s simply unhealthy evaluation.
Alex Chen Alex Chen Read More
