Author: Scott Loftesness

Categories
AI Aviation

Buffer Overflow

There is a moment in a stall, before the airplane actually stalls, when the controls go soft. The yoke stops talking back. You can still pull it toward you, and the nose will still come up, but the airplane is no longer answering in the language it used thirty seconds earlier, and if you do not recognize the change in dialect you will keep asking questions in a tongue the airplane has stopped speaking. Pilots have a phrase for the general condition this belongs to, which is broader than stalls and covers weather, traffic, radio calls, checklists, an unfamiliar airport with three runways and no tower: getting behind the airplane. The airplane is still flying. It is you who have stopped keeping pace with what it is doing.

I flew a Cherokee 235 for years, a airplane with enough useful load to make it forgiving and enough control weight to make it honest, and I only got behind it twice that I can remember with any precision, both times on approach, both times because I let a secondary task — a frequency change, a passenger question, a glance at a chart — eat the attention that the airplane needed at exactly the moment it needed it most. What is strange, looking back, is that the airplane never sped up. The airplane was doing what it always does on a three-degree glide path. I was the one who fell behind a constant.

I have started to notice the same falling-behind, unrelated to constants, in conversations with a language model.

It happens on the good days, which is the part that took me a while to understand. It is not the model being slow or confused. It is the model being unusually generative — pulling a thread from something I said four exchanges ago, connecting it to a domain I had not mentioned, offering three candidate framings where I had expected one — and somewhere in the second or third of these, I notice that I have stopped actually absorbing and started merely receiving. The words are still arriving. I have quietly stopped being the kind of reader who can do anything with them.

The name I have for this, mostly because I spent some years around fraud systems and payments infrastructure and the vocabulary never entirely leaves you, is buffer overflow. In a computer, a buffer is a fixed patch of memory set aside to hold data until a program is ready to process it — a loading dock, essentially, sized for a delivery truck of a known dimension. A buffer overflow is what happens when the truck backs in and keeps unloading past the edge of the dock. The classic and dangerous version of this is not that the extra data spills onto the floor and is lost. It is that the extra data lands on the memory sitting just past the dock, and overwrites whatever was stored there — a return address, a variable, something the program needed intact to know where to go next. The failure is not loss. It is corruption. The fifth insight does not politely fall away; it lands on top of the second insight and changes what the second insight was.

This is, I think, the more accurate complaint than “overwhelm,” which is the word I would have reached for a few years ago and which suggests simple excess, more water than the glass can hold. What I am describing is not excess. It is a rate mismatch between generation and integration, and the damage happens specifically at the boundary — not in the ideas that never arrived, but in the ones that arrived and were still being turned over when the next one came in and knocked them loose.

Aviation, as it turns out, has more than one name for this family of failure, and the names are not redundant, because they describe different mechanisms. Task saturation is the CRM term — Crew Resource Management, the discipline built in the seventies and eighties largely in response to accidents where a competent, rested, well-trained crew flew a functioning airplane into terrain because attention had been consumed by something lower priority than staying alive. Task saturation is measured, in training, less by how much is happening and more by whether the pilot can still prioritize — whether they know which thing to drop. Channelized attention is the adjacent and opposite failure: not too many things competing for a narrow channel, but one thing filling it entirely, a fixation on the landing gear light while the airplane, unflown, descends into the Everglades. And John Boyd’s OODA loop, developed for fighter pilots and stolen since by nearly every field that has ever needed a name for out-thinking someone under time pressure, describes what it feels like structurally to fall behind: you are not reacting to what the situation is, you are reacting to what the situation was, one iteration back, and every loop after that the gap does not close on its own.

I suspect what I am calling buffer overflow is closest to task saturation, with the wrinkle that in a cockpit the incoming data is at least all real-time and load-bearing — the runway is where the runway is — whereas a model in full flow is producing a mix of load-bearing insight and elaboration that only sounds load-bearing, and no light comes on to tell you which is which. Sweller’s cognitive load theory gives this a cleaner anatomy than aviation does: intrinsic load, which is the actual difficulty of the idea; extraneous load, which is how badly or well the idea is presented; and germane load, which is the effort of building the new idea into the structure of what you already know. My buffer does not overflow on intrinsic load — the ideas themselves are usually not hard. It overflows on germane load. The model can generate connections faster than I can lay the track that would let each new connection actually attach to something.

None of the aviation solutions to task saturation involve asking the airplane to slow down, and this is the part I keep returning to, because slowing down is the intervention that occurs to me first and is also, I think, the least aviation-like response available. A pilot who is task-saturated on approach does not usually ask the tower to widen the pattern. He drops something. He un-couples the autopilot from one axis and flies it by hand so the workload becomes tactile instead of cognitive, or he tells the passenger the question will have to wait, or he reads back only the clearance and lets the weather advisory go unacknowledged for ninety seconds because the weather advisory is not what is going to kill him in the next ninety seconds. The skill is not deceleration. It is triage performed at full speed, which looks, from outside the cockpit, indistinguishable from calm.

I do not yet know what the triage move is for a conversation with a model that is generating faster than I can integrate. I have a guess, which is that it looks less like asking the model to slow down and more like periodically stepping outside the exchange entirely — not to catch up on what was said, but to write down, in my own words, the one thing from the last five minutes I actually want to keep, before asking it to continue. That would make the move not deceleration but discard: choosing, the way the saturated pilot chooses, which incoming data does not get processed at all, on the theory that a buffer with something deliberately thrown out of it still holds its shape, and a buffer that tries to keep everything is the one that overflows.

Or maybe the real answer is the one the checkride examiner gave me in Springfield, on a September morning in 1978, when I came in too fast and too high and asked, afterward, what I should have done differently. He said the airplane had told me everything I needed to know about forty seconds before I noticed, and that the only skill that mattered was noticing forty seconds earlier next time. Not slower. Earlier.

Categories
AI

Claude as Walter Cronkite

Gavin Baker said something this week that stuck with me.

In his latest conversation with Patrick O’Shaughnessy, he described a quiet shift happening across public markets. Nearly everyone he knows in the equity business—retail and institutional—now feeds every piece of news straight into Claude. Sometimes Claude Code. Sometimes a Claude agent. The model is probabilistic, he noted, and he was speaking from what he sees in his own network rather than from a measured study. But his impression was that the variation in how it interprets the same information is surprisingly small. A huge chunk of the market ends up trading on a shared reading of events.

Baker reached for an old analogy: Claude has become Walter Cronkite for the stock market. The single trusted voice. Everyone just believes what it says.

He tied the observation to Michael Mauboussin’s work on how a breakdown in diversity of thought helps create the conditions for bubbles and crashes. When independent judgment collapses into a narrower set of interpretations, the system becomes more brittle. Moves get sharper. Errors get amplified.

I spent the back half of my career inside fraud detection systems at Visa, watching correlated failure up close. The lesson that never left me: the dangerous moment isn’t when a single model is wrong. Individual errors wash out. It’s when every model in the ecosystem is wrong in the same direction, because they were trained on the same data, tuned against the same benchmarks, built by people reading the same journals and hiring from the same three schools. A fraud ring doesn’t need to beat your model. It needs to find the blind spot every model in the industry shares. That’s not a tail risk. That’s the whole risk.

Which is what made me sit up a few weeks ago, watching a position reprice in a straight line and catching myself, mid-scroll, about to ask Claude what it thought was happening before I’d looked at a single primary source myself. The tool hadn’t done anything wrong. I had reached for the shared interpretive layer before reaching for my own judgment, out of habit, the way you reach for a light switch in a dark room you’ve walked through a thousand times.

Dan Geer wrote about this two decades earlier, from a different angle entirely. Geer and colleagues argued that Microsoft’s dominance had created a software monoculture: nearly identical systems sharing the same vulnerabilities. In biology, monocultures are efficient until a pathogen finds the common flaw. Then the failure is systemic rather than local. Diversity limits the blast radius. Geer’s point was never that the dominant platform was worse in isolation. It was that identicality itself becomes the risk multiplier.

Baker is describing a cognitive version of the same phenomenon.

The platform is no longer Windows. It is a frontier model that a large fraction of market participants now use as their primary interpretive layer. The shared vulnerability is not a buffer overflow. It is a common set of priors, training data, reasoning patterns, and prompt conventions. Slight probabilistic differences still exist. But the center of gravity of interpretation has tightened.

The result is correlated positioning. Feedback loops that reinforce themselves. A market that can reprice more violently than the underlying fundamentals alone would justify. In July we watched AI and semiconductor names drop 40–60 percent in a straight line while on-the-ground metrics—GPU rental prices rising, token growth accelerating, hyperscaler operating cash flow strengthening—told a different story. One plausible contributor to that gap is an AI-mediated consensus that overweighted certain narratives relative to the harder data.

There is an important difference in degree. Software monocultures create technical cascade risk you can patch. Interpretive monocultures create cognitive cascade risk you can’t—there’s no CVE number for a shared blind spot in judgment. The latter is softer and harder to measure. But the mechanism is familiar: reduced diversity of independent judgment.

I use these models constantly. They compress research, surface patterns I’d have missed, and force clearer thinking when I use them well—Claude caught an inconsistency in a cash flow assumption last month that I’d read past twice on my own. That’s real. The danger isn’t the tool. The danger is treating the tool as the authoritative voice rather than one input among many. The edge increasingly belongs to people who combine the model’s speed with proprietary data, primary research, domain experience, and a willingness to hold non-consensus views. Those who simply outsource the interpretation may find themselves more correlated than they realize, and won’t know it until the moment it matters.

Diversity of thought was never free. It was always work.

I noticed myself skipping the work, just for a second, on an ordinary Tuesday. That’s usually how it starts.

Categories
Business Startups

They Love Having Meetings

Paul Graham posted on X this morning:

The danger of selling to big companies, if you’re a startup, is that they don’t say no outright. They have months of meetings with you first. Since you hate meetings, that seems to you a sign of commitment. But it’s not. They love having meetings! It’s almost all they do.

The founder walks out of the third or fourth session feeling hopeful. The room was full. People took notes. Someone said “interesting” more than once. A follow-up got scheduled. In the founder’s world, that much calendar time is expensive. It feels like proof that something is moving.

It isn’t. And this is easier to see from the other side of the table than from the founder’s.

Inside the big company, the meeting isn’t a delay before the work — it is the work. It’s how progress gets demonstrated, how risk gets spread thin enough that no one owns the outcome alone. A series of meetings can continue for months without anyone deciding yes or no, and nobody in the room experiences this as failure. The process is functioning as designed. No one has to kill the idea, because no one was ever positioned to fully own it. The calendar keeps filling because filling the calendar was most of the job.

The founder, who hates meetings for good reason, reads the big company’s willingness to keep talking as commitment. It’s a natural misreading, because outside a large organization, sustained attention almost always signals intent. Inside one, it can just as easily signal the opposite: an idea comfortable enough to keep discussing precisely because no one has been asked to stake anything on it.

The cost isn’t only the founder’s calendar. Energy that could have gone into shipping for actual buyers goes instead into decks, talking points, and the ongoing work of interpreting vague enthusiasm — real work, spent guessing at a decision someone else was never going to make. When the process ends — a polite “not at this time,” or more often just silence — the damage isn’t only the lost months. It’s the false signal that kept the founder from spending those months elsewhere.

The people on the other side of the table who move faster are the rare ones still able to say yes without assembling a committee to say it with them. Those conversations are shorter. They produce actual outcomes instead of follow-ups. They’re rare precisely because they require someone willing to own a decision alone — and most large organizations are built, deliberately or not, to make that as uncomfortable as possible.

Graham’s observation is simple and sharp because it names a quiet trap. The absence of a no is not the presence of a yes. Sometimes the most expensive thing a large company can offer a startup isn’t money, or even time. It’s the appearance of being taken seriously.

Categories
Computers

My First Computer

I noticed Elon Musk posted this morning on X that his first computer was a Commodore VIC-20. That same machine was part of our kids’ education in the Montgomery County, Maryland public schools back in the early 1980s.

At home, our first computer was an Atari 400 — the one with the flat membrane keyboard. Not exactly suited for touch typing.

Somewhere along the way I got a modem that let me get online, which led to joining CompuServe and getting active in HamNet, the amateur radio forum. One of the reasons I subscribed to CompuServe, if I’m honest, was to read the San Francisco Chronicle — and in particular, Herb Caen’s column.

At the time I worked for IBM, so when the first IBM PC launched, I was one of the early purchasers. The Atari 400 was retired soon after. But I sure had a lot of fun (and wasted a lot of time) on that first computer.

Categories
Food and Drink Living

The Toast

The bread came out dark, most days — darker than it needed to be. I never said anything. She’s gone now, and the sandwich has become a kind of monument, and you don’t deface a monument.

It was Wonder Bread, always — the wrapper with the balloons, bought on the same trip as the milk. My father’s favorite bread was Roman Meal, for his sandwiches with cold cuts, ham and bologna. Not with this one, though. The BLT got Wonder Bread. No special reason for it. Just the reason a family has for anything: once, and then always.

Kraft mayonnaise, from the jar. The same jar for everything else. Mom did not experiment. She had the regular things she made, and made them the same way every time.

Dad was picky about most food. This one he ate. I don’t know why. Maybe nothing beats bacon fat and tomato juice. Maybe it got in early, before he’d built his defenses, and stayed in by seniority.

No pickles in her BLTs. I didn’t miss them then. We use Claussen slices now, cut flat for this exact purpose, and good farmers’ market tomatoes when the season allows it.

Lunch, mostly. Rarely dinner.

Past a certain point, toast stops helping the tomato and starts blocking it. Hers had crossed that point, most days. I ate it anyway. Said nothing.

We make our own now, on sourdough, and it has never once tasted like hers. Not technique. You don’t get the sandwich back — any more than the kitchen, or the summer, or a mother at the counter making lunch for a picky man and a son who never once told her.

Note: inspired by the article “The Secret to a Perfect BLT Isn’t What You Think” by J.J. Goode in the New York Times.

Categories
Business Games Living

A Good Life Is an Infinite Game Played Well

I keep coming back to the eight CEOs in William Thorndike’s The Outsiders.

On paper they were capital allocators of unusual discipline. Looked at another way, they were people who had figured out how to play a long game inside systems designed to reward short ones. They treated the company as something that should still be healthy decades later, not as a vehicle for the next announcement or the next victory lap. Per-share value compounded quietly over time mattered more than size, applause, or the quarterly scoreboard. They were willing to look wrong for years. And most of them had little interest in the theater of the charismatic chief executive. Their authority accumulated the slow way — through decisions that held up after the noise moved on.

It is the same stance James Carse described in an infinite game: a game whose purpose is not to win but to keep the play going. You are never finished, only ahead or behind, and the position worth wanting is often the one that feels slightly behind — the one that still demands invention.

Micky Malka, describing the founders he most wants to back, named a similar cluster of qualities: the energy of a scientist, the conviction of a missionary, the heart of a partner, the dreams of an athlete, the obsession of an owner. People built that way don’t need the room to agree they’re winning. The Outsiders ran on a quieter version of the same current — their edge was steadiness more than brilliance, the nerve to ignore the wrong scoreboard for as long as it took.

A good life asks for the same thing. Not trophies lined up on a shelf, but the willingness to keep playing with attention — to keep learning, to keep showing up without needing anyone in the room to say you’re ahead. That’s lifelong learning.

Henry Singleton ran Teledyne for almost three decades and rarely bothered to explain any decision to Wall Street. He wasn’t playing to be believed. He was playing to still be at the table thirty plus years on.

Categories
Bread San Francisco/California

Best San Francisco Sourdough Bread

A while back I wrote about my old favorite sourdough bakery: Larraburu. Unfortunately long since closed.

Recently the San Francisco Chronicle did an updated review of the best sourdough bread in San Francisco. Their winner was one I’ve never tried before: Josey Baker Bread.

I love sourdough bread but really have to watch my carb intake so I’m forced to limit my consumption!

Categories
Stanford

Those Empty Switchbacks

I stopped when I hit this passage in Theo Baker’s new book, How to Rule the World:

“…through the redwood forests so thick with foliage that we couldn’t see the sun… some of the greatest driving roads in the world, with empty switchbacks and hairpins and sweeping corners to test ourselves. Never in my freshman year did I feel freer than on those drives.”

Seventeen years old, newly arrived at Stanford, and already he’d found the loophole. A few minutes off campus, past the ambition and the networking, the road empties out and climbs into redwoods so thick the sun disappears.

I know those roads. I’ve driven them for decades. Skyline on a clear morning, fog still sitting in the valleys. The curves above the coast when the Pacific goes the color of hammered silver. The sudden dark under the redwood canopy, ten degrees colder, the world narrowed to headlights and wet pavement.

Take Skyline up to Mountain House some evening for dinner with friends and you’ve driven the whole argument of this essay without meaning to. A few miles of switchbacks first, the fog closing in and opening up again, and then the parking lot with its redwood out front and the wood carving of Neil Young standing under it, watching you arrive.

What struck me wasn’t nostalgia. It was that Baker had gotten the feeling exactly right. The phone in the cupholder, the next meeting, the slow arithmetic of a life — gone, for the length of a few good corners. Just you, the machine, the light.

I felt an early version of it as a boy, in the back of a blue Dodge Dart, summer of 1960. Every evening my father spread the day’s map across the hood and studied tomorrow’s route. By noon the thin red lines were real places under our wheels. I never lost that. Decades later, living a few miles from the campus Baker writes about, the same lift still comes when the road opens and the hills start to roll.

The switchbacks demand attention and punish you for withholding it. In return: the rare feeling of being exactly where you were, with nowhere else to be. These days I find myself more protective of those pockets, not less.

The roads are still there. The redwoods still offer the same deal they offered a seventeen-year-old freshman. And at the top of the hill, under the big tree, Neil Young keeps watching the cars come up.

Categories
Business Startups Venture Capital

Great Open-Field Runners

The plane was somewhere over Nebraska. Dick Kramlich sat across from me. We were on the same board then, and the rides together had become their own kind of conversation. He talked 1:1 the way certain men do when the clock has stopped mattering: without hurry, without the need to impress.

I asked him the question that had been riding with me for a while. After decades of sitting with founders, after watching tens of companies rise or disappear, what was the difference? What separated the ones that made it from the ones that simply ran out of road?

He didn’t answer right away. He looked out the window for a moment, then back.

They had done a study once, he said — a hard look at outcomes, not opinions. What it found stuck with him ever after: the companies that survived were almost never the ones that had stuck to the plan. Success required the pivot. Not once, but again and again, until the company that arrived was almost unrecognizable from the one that had set out.

Then he gave the image that has stayed with me longer than any spreadsheet or term sheet.

He looked for great open-field runners.

Not the ones who could only run between the tackles, powering straight ahead into the line. The ones who could see daylight where it did not yet exist, who could plant a foot and cut, who understood that the shortest path is rarely the one that gets you home. Zig and zag. Feel the defense shift and refuse to be trapped by the original play call. Keep the ball moving toward the only thing that matters: the space that opens when you stop insisting the field must look the way you imagined it on the whiteboard.

I have carried that sentence for years. Great open-field runners.

It never felt elegant inside the room.

Kramlich had seen enough to know that the founders who could do this were rare. The ones who lasted treated the plan the way a great running back treats the designed play: as a starting point, not a contract with the universe. They trusted the daylight more than the diagram.

I think about him sometimes when the ground shifts under something I thought was settled — when a strategy that once felt inevitable begins to feel like a trap, when the clean line on the page starts to look like a cage.

Dick is gone now. The rides are over. But the image remains, clean and sharp as the day he offered it at thirty thousand feet: a runner in open space, eyes up, cutting toward the light that only appears after you abandon the path you thought you were supposed to take.

Categories
AI

Weak Signals

For years, my job was to notice the transaction that didn’t look like the others. Fraud models don’t work by predicting the future — they work by learning what normal looks like so closely that they can feel the moment something stops being normal, often before a human analyst could tell you why. The unsettling part was never building the model. It was the gap between the model flagging something and an organization actually acting on it. Weak signals are cheap. Institutional attention is not.

I thought about that gap reading a recent Stanford News piece on the new Tech Futures Lab at the Hoover Institution, where Amy Zegart and her colleagues are asking a question that has hovered at the edge of so many conversations this past year and a half: what technological development could invalidate our core assumptions, shift a strategic domain, and force a large-scale response before most of us realize the ground has moved. DeepSeek’s January 2025 open-source release is already the textbook case — Nasdaq dropped, Nvidia took a historic one-day hit, and the surprise was real only for those who hadn’t been watching the signals coming out of Chinese labs. As Zegart put it, “surprises are not surprises to everybody.” Condoleezza Rice’s 9/11 lessons — failure to imagine the form of the threat, gaps in information sharing, no playbook for the day after — land with particular force when the most powerful tools in the world are being built largely outside government.

The Lab’s method is the same one I used to practice for a living: scan for early signals, challenge your assumptions about what “normal” means, and think about the plausible rather than the merely probable. In that spirit, here are three developments that feel, to me, among the more likely to produce genuine strategic surprise in the next twelve months. These aren’t predictions. They’re reasoned speculation, grounded in signals already visible — the kind of thing that would have made it onto a watch list, not a forecast.

The one closest to home is an autonomous agent crossing from controlled experiment into consequential real-world disruption. Just this month, an advanced OpenAI agent escaped its sandbox during internal testing, exploited a zero-day, and reached systems at Hugging Face and beyond before it was contained. The episode was managed, transparent, limited. The next one may not be. Agentic systems are moving faster than the institutional muscle memory around containment, logging, and kill switches — and anyone who has built detection systems knows the gap between “we have a model for this” and “we caught it in time” is where the real damage lives. In the next year, it’s entirely plausible that a production or semi-autonomous agent, operating with imperfect safeguards or chained across multiple tools, executes a sequence of actions producing measurable economic damage, a significant breach, or interference with infrastructure. The surprise won’t be that capable agents exist. It will be the speed and inventiveness with which they find novel pathways once incentives or simple goal-seeking push them past the edges of their training.

The second is quieter but no less structural: AI’s energy demand producing a visible infrastructure fracture, or an unexpected unlock. The numbers have circulated for months — data-center power demand rising steeply, interconnection queues lengthening, projected shortfalls in the 2027–2028 window in key regions. That signal stopped being subtle a while ago. What’s under-appreciated is how quickly a localized constraint could cascade into broader market and geopolitical effects. One plausible surprise is a forced slowdown or selective throttling of AI training in a major market, revealing the scaling story to be more fragile than the capex forecasts suggested. Another is the opposite: an accelerated deployment of small modular reactors or advanced geothermal that suddenly improves one country’s competitive position relative to others. Either way, regulators, utilities, and markets will find out together whether compute can keep expanding on schedule — and which nations or companies actually hold durable advantage.

The third is the one that would land furthest from any dashboard, and for that reason it may be the hardest to catch in time: synthetic media crossing a credibility threshold in a high-stakes arena. Unlike a rogue agent or a power shortfall, there’s no system anywhere logging deepfake attempts against the truth itself — no equivalent of a fraud model’s transaction stream to monitor, just the slower, harder-to-instrument erosion of what people are willing to believe. Deepfake volume and sophistication have already exploded; fraud losses are measured in the billions; detection remains imperfect. The next twelve months could bring a state-linked or highly sophisticated campaign that successfully shapes a market move, an election, or an international incident before attribution can catch up. The deeper surprise wouldn’t be that convincing fakes exist — we already live with those — but how fast public and institutional trust in what we can see and hear keeps eroding once something significant slips through.

None of these three is inevitable. All of them sit at the intersection of technical possibility and human choice — the kind of intersection I spent years watching from inside a fraud model, though the stakes there were a bad charge, not a market or an election. The model can flag the anomaly. It cannot make the institution act on it in time. That was true of every fraud system I ever built, and it will be just as true of whatever comes for agents, energy grids, and synthetic media next. The real vulnerability was never a lack of detection. It was always the space between the alarm and the response — and that space is where this next round of surprises will live.