Camille Tomlin, VP of IT systems and infrastructure at Philadelphia International Airport, ends up arguing that civic infrastructure works when nobody notices it, and that public-sector IT is a deliberate career choice offering a breadth of technology private jobs rarely do.
The landlord model. Tomlin frames the airport as landlord to airlines and tenants: her team runs base infrastructure and fiber, extending a carrier's circuit to its closet and handing off, as it did for Aeromexico in June. American Airlines, with PHL as a hub, manages its own terminals. "If you've been to one airport, you've been to one airport."
A mini city on cable. She describes the airport as a city with fire, police and federal agencies, plus FAA-regulated aviation, and says anything with a chip that needs a cable involves her team. The FAA tower is separate from ground towers. On security, prompted by Hanselman's Chromecast anecdote, she cites closed on-prem networks, enterprise displays, an infosec team and pen tests, and urges travellers to verify the SSID on the airport's website.
Quiet wins and constraints. Her proudest result is uptime and redundancy passengers never see. Capital projects keep adding sensors, and she wants to apply AI to that data, but each addition must pass the test of sustainable opex under government budget constraints: "slightly behind but not 10 years."
Public service as a career. Tomlin traces her path from Drexel co-op and web development to IT director at the Free Library of Philadelphia, then to the airport, with a team of a little under 80. Airport IT leaders compare notes through Airports Council International. She argues city jobs are publicly posted, that curiosity matters more than technical depth as one moves into management, and that policymakers should back the dedicated people who maintain these services.
On a quiet, rally-heavy pre-Thanksgiving session -- the S&P and NASDAQ both roughly 1% from all-time highs with almost no sector rotation -- the hosts spend the day on a single connecting idea: whether it's a company deciding what to do with excess cash, a CEO deciding what to say publicly about their own stock, or an individual deciding whether to take risk at work, the honest answer usually differs from the convenient one.
Borrowing, volatility, and career mailbag. Asked whether to borrow against a house or against Bitcoin, the hosts land on rate-shopping over asset-class preference, though one favors the house simply because it carries far less volatility as collateral. A question about whether market volatility affects hiring gets a nuanced answer: day-to-day swings mean little, but the tariff-driven volatility since April has visibly made companies more cautious about hiring, and the hosts note their new platform's live jobs-data feed shows stock market performance and interest rates are measurably built into how companies price roles. On job-hopping, both hosts agree that leaving positions within six to twelve months is close to a disqualifying red flag in hiring -- it takes at least six months to actually learn a role, and losing a trained employee is among the costliest mistakes a business can make -- while movement within a single company via promotion is viewed completely differently. A post-MBA founder asking how to raise capital with minimal experience is told the honest constraint is a thin network and short track record, illustrated by one host's own first capital raise ($10 million for a trading fund in the late 1980s) coming only after seven or eight years of visible success; the deeper point is that raising money requires being comfortable asking and absorbing repeated rejection. A Fortune 500 employee weighing whether to take risk at work is told corporate risk-reward is fundamentally asymmetric compared to being an owner -- credit for a good bet gets diluted moving up a hierarchy, so the likelier payoff for taking career risk is a better opportunity elsewhere rather than a reward at the current job.
Stock buybacks, weighed honestly. The main segment runs through the tradeoffs between the four uses of excess corporate cash -- buybacks, acquisitions, dividends, and reinvestment -- with both hosts, who've sat on both sides of public-company buyback decisions, expressing real ambivalence. Buybacks read as a "safe play" that signals confidence to the market, but the hosts cite IBM's roughly $150 billion in buybacks over two decades against a company worth about $170 billion as proof the signal doesn't guarantee a good outcome. Dividends draw more consistent skepticism from one host, who'd always rather see cash redeployed to grow the business than paid out. Acquisitions get framed candidly: buying an innovative smaller company is usually cheaper than trying to build that innovation internally, which is why premiums for strategic acquisitions are typically justified rather than overpaying, and "fire sale" bargains are rare. On the downside, both agree buybacks can divert money from R&D and wages, disproportionately benefit already-wealthy shareholders, and inflate the same stock-price metrics executive compensation is tied to -- a dynamic one host compares to a sports coach who's effectively "hired to be fired" within a matter of years.
CEOs pumping their own stock. A scripted rant argues that Nvidia's Jensen Huang and AMD's Lisa Su made strongly bullish public statements about AI demand and valuation even as their own SEC filings show substantial personal stock sales -- roughly $13 million a day for Huang since summer, $37 million for Su -- contrasting them with executives like Buffett, Musk, or Jamie Dimon who have publicly acknowledged their own companies might be overvalued. The argument isn't that the CEOs are lying, but that executive compensation tied to share price structurally rewards optimistic public messaging regardless of the underlying risk to retail investors who take the "buy" signal at face value; both Nvidia and AMD shares fell meaningfully in the weeks following the comments cited.
Career pivots and the platform's pitch. The closing segment previews a Lost Dog feature that maps a user's stated skills against seven industries -- finance, education, healthcare, government, retail/consumer, tech, and business services -- to suggest where a given skill set might transfer best, alongside the now-familiar advice to self-assess honestly, start any pivot small, and keep learning continuously.
On a quiet, rally-heavy pre-Thanksgiving session -- the S&P and NASDAQ both roughly 1% from all-time highs with almost no sector rotation -- the hosts spend the day on a single connecting idea: whether it's a company deciding what to do with excess cash, a CEO deciding what to say publicly about their own stock, or an individual deciding whether to take risk at work, the honest answer usually differs from the convenient one.
Borrowing, volatility, and career mailbag. Asked whether to borrow against a house or against Bitcoin, the hosts land on rate-shopping over asset-class preference, though one favors the house simply because it carries far less volatility as collateral. A question about whether market volatility affects hiring gets a nuanced answer: day-to-day swings mean little, but the tariff-driven volatility since April has visibly made companies more cautious about hiring, and the hosts note their new platform's live jobs-data feed shows stock market performance and interest rates are measurably built into how companies price roles. On job-hopping, both hosts agree that leaving positions within six to twelve months is close to a disqualifying red flag in hiring -- it takes at least six months to actually learn a role, and losing a trained employee is among the costliest mistakes a business can make -- while movement within a single company via promotion is viewed completely differently. A post-MBA founder asking how to raise capital with minimal experience is told the honest constraint is a thin network and short track record, illustrated by one host's own first capital raise ($10 million for a trading fund in the late 1980s) coming only after seven or eight years of visible success; the deeper point is that raising money requires being comfortable asking and absorbing repeated rejection. A Fortune 500 employee weighing whether to take risk at work is told corporate risk-reward is fundamentally asymmetric compared to being an owner -- credit for a good bet gets diluted moving up a hierarchy, so the likelier payoff for taking career risk is a better opportunity elsewhere rather than a reward at the current job.
Stock buybacks, weighed honestly. The main segment runs through the tradeoffs between the four uses of excess corporate cash -- buybacks, acquisitions, dividends, and reinvestment -- with both hosts, who've sat on both sides of public-company buyback decisions, expressing real ambivalence. Buybacks read as a "safe play" that signals confidence to the market, but the hosts cite IBM's roughly $150 billion in buybacks over two decades against a company worth about $170 billion as proof the signal doesn't guarantee a good outcome. Dividends draw more consistent skepticism from one host, who'd always rather see cash redeployed to grow the business than paid out. Acquisitions get framed candidly: buying an innovative smaller company is usually cheaper than trying to build that innovation internally, which is why premiums for strategic acquisitions are typically justified rather than overpaying, and "fire sale" bargains are rare. On the downside, both agree buybacks can divert money from R&D and wages, disproportionately benefit already-wealthy shareholders, and inflate the same stock-price metrics executive compensation is tied to -- a dynamic one host compares to a sports coach who's effectively "hired to be fired" within a matter of years.
CEOs pumping their own stock. A scripted rant argues that Nvidia's Jensen Huang and AMD's Lisa Su made strongly bullish public statements about AI demand and valuation even as their own SEC filings show substantial personal stock sales -- roughly $13 million a day for Huang since summer, $37 million for Su -- contrasting them with executives like Buffett, Musk, or Jamie Dimon who have publicly acknowledged their own companies might be overvalued. The argument isn't that the CEOs are lying, but that executive compensation tied to share price structurally rewards optimistic public messaging regardless of the underlying risk to retail investors who take the "buy" signal at face value; both Nvidia and AMD shares fell meaningfully in the weeks following the comments cited.
Career pivots and the platform's pitch. The closing segment previews a Lost Dog feature that maps a user's stated skills against seven industries -- finance, education, healthcare, government, retail/consumer, tech, and business services -- to suggest where a given skill set might transfer best, alongside the now-familiar advice to self-assess honestly, start any pivot small, and keep learning continuously.
wsl --install and following the prompts for a username and password installs Ubuntu by default; wsl --list --online shows the alternatives, such as Debian, Kali and Alpine, and typing wsl or launching Ubuntu from Start gets back in.sudo apt update && sudo apt upgrade -y.Trading through a genuinely volatile session -- the NASDAQ swings from down roughly 400 points to unchanged while Nvidia, AMD, and Oracle sell off even as the broader market rallies -- the hosts spend the day defending the idea that big, hard-to-quantify decisions (going to college, choosing a mortgage term, taking feedback) deserve options and honesty rather than paternalistic rules imposed by people who think they know better.
Mailbag and the Lost Dog origin story. A consultant asking whether the coming Lost Dog software can price hourly or daily rates is told the platform works in annual terms, but dividing an annual figure by 2,000 hours gets a rough answer; the exchange turns into a brief aside on exempt-versus-non-exempt labor law and why the hosts have never structured their own hires around hourly billing. Asked whether the show will keep discussing markets and trading, the hosts confirm yes, but frame it less as active-trading instruction and more as building financial literacy and comfort with two-sided markets, tying it to a coming portfolio feature on the platform. A charming aside reveals the platform's name traces to a "Lost Cat" poster spotted in an Atlanta green room years earlier -- the hosts liked the format enough to build a "Lost Dog" counterpart, distinct from their prior brand names, which were simply invented on the spot. A 20-year-old engineering student is told the software will let him model his profile against any profession or city, not just his intended path, functioning as a two-sided market for a career decision rather than a verdict on which major to choose.
Should you go to college? The day's central segment weighs the cost of a degree (roughly $27,000 a year in public tuition, $58,000 private, $30,000-$40,000 in average student debt) against $1.2 million in additional lifetime earnings and roughly half the unemployment rate of high school graduates, alongside trade school and coding bootcamps as faster, cheaper alternatives for people without the option or interest in a four-year path. The hosts largely agree the numbers favor college for anyone who can afford it, but split on why: one frames the choice as maximizing return on a specific career goal, while the other flatly rejects that framing, using his own path -- a political science degree, unrelated internships, and a Wall Street job offer landed on confidence and networking rather than subject expertise -- as proof that college's real value is the maturation, friendships, and confidence gained along the way, regardless of major. Both agree a résumé line about where someone went to school matters mainly in the first few years of a career and fades after that, though one host concedes that categorically refusing to consider candidates without a degree risks eliminating people who built comparable skill through experience instead.
A 50-year mortgage, and growth mindset. A scripted rant argues financial commentators criticizing the newly floated 50-year mortgage option are missing that the realistic alternative for many buyers isn't a standard 30-year loan but no homeownership at all, framing the roughly $226 monthly payment reduction as meaningful for buyers otherwise locked out, and casting resistance to the option as a kind of financial paternalism inconsistent with how leverage is already extended to wealthier buyers. The closing segment on growth mindset surfaces a running disagreement over terminology -- one host prefers to call it "positive drift," the constant, cumulative process of getting better, while the other insists growth mindset specifically includes learning from failure -- and a sharper split over feedback itself: one host calls "constructive criticism" inherently passive-aggressive, preferring blunt direct correction, while the other considers gently naming what isn't working to be constructive by definition, calling the disagreement mostly semantic. Both agree, without disagreement, that hard work in one's younger years pays off later even at some cost to work-life balance, illustrated by a story about a trading-desk mistake during COVID that could have ended someone's job but instead became, in the hosts' telling, a formative lesson survived rather than a career-ending failure.
Trading through a genuinely volatile session -- the NASDAQ swings from down roughly 400 points to unchanged while Nvidia, AMD, and Oracle sell off even as the broader market rallies -- the hosts spend the day defending the idea that big, hard-to-quantify decisions (going to college, choosing a mortgage term, taking feedback) deserve options and honesty rather than paternalistic rules imposed by people who think they know better.
Mailbag and the Lost Dog origin story. A consultant asking whether the coming Lost Dog software can price hourly or daily rates is told the platform works in annual terms, but dividing an annual figure by 2,000 hours gets a rough answer; the exchange turns into a brief aside on exempt-versus-non-exempt labor law and why the hosts have never structured their own hires around hourly billing. Asked whether the show will keep discussing markets and trading, the hosts confirm yes, but frame it less as active-trading instruction and more as building financial literacy and comfort with two-sided markets, tying it to a coming portfolio feature on the platform. A charming aside reveals the platform's name traces to a "Lost Cat" poster spotted in an Atlanta green room years earlier -- the hosts liked the format enough to build a "Lost Dog" counterpart, distinct from their prior brand names, which were simply invented on the spot. A 20-year-old engineering student is told the software will let him model his profile against any profession or city, not just his intended path, functioning as a two-sided market for a career decision rather than a verdict on which major to choose.
Should you go to college? The day's central segment weighs the cost of a degree (roughly $27,000 a year in public tuition, $58,000 private, $30,000-$40,000 in average student debt) against $1.2 million in additional lifetime earnings and roughly half the unemployment rate of high school graduates, alongside trade school and coding bootcamps as faster, cheaper alternatives for people without the option or interest in a four-year path. The hosts largely agree the numbers favor college for anyone who can afford it, but split on why: one frames the choice as maximizing return on a specific career goal, while the other flatly rejects that framing, using his own path -- a political science degree, unrelated internships, and a Wall Street job offer landed on confidence and networking rather than subject expertise -- as proof that college's real value is the maturation, friendships, and confidence gained along the way, regardless of major. Both agree a résumé line about where someone went to school matters mainly in the first few years of a career and fades after that, though one host concedes that categorically refusing to consider candidates without a degree risks eliminating people who built comparable skill through experience instead.
A 50-year mortgage, and growth mindset. A scripted rant argues financial commentators criticizing the newly floated 50-year mortgage option are missing that the realistic alternative for many buyers isn't a standard 30-year loan but no homeownership at all, framing the roughly $226 monthly payment reduction as meaningful for buyers otherwise locked out, and casting resistance to the option as a kind of financial paternalism inconsistent with how leverage is already extended to wealthier buyers. The closing segment on growth mindset surfaces a running disagreement over terminology -- one host prefers to call it "positive drift," the constant, cumulative process of getting better, while the other insists growth mindset specifically includes learning from failure -- and a sharper split over feedback itself: one host calls "constructive criticism" inherently passive-aggressive, preferring blunt direct correction, while the other considers gently naming what isn't working to be constructive by definition, calling the disagreement mostly semantic. Both agree, without disagreement, that hard work in one's younger years pays off later even at some cost to work-life balance, illustrated by a story about a trading-desk mistake during COVID that could have ended someone's job but instead became, in the hosts' telling, a formative lesson survived rather than a career-ending failure.
Maya Ackerman, associate professor of computer science and engineering at Santa Clara University and CEO and co-founder of Wave AI, ends where her new book Creative Machines starts: machines are already creative, humans are nowhere near the ceiling of their own creativity, and generative AI should be built to elevate people rather than act as an all-knowing oracle that replaces them.
Machines are creative, and have been for a while. Ackerman rejects both the sci-fi story of replacement and the denial that AI is creative. Her evidence is David Cope's Experiments in Musical Intelligence, which in the 1980s composed Bach- and Vivaldi-style music listeners could not tell from the originals. Industry arrived only about ten years ago; convincing generated sound, as Suno and Udio produce, dates to about 2022. She credits the current leap to money for bigger models and marketing, not new ideas, and argues that because we are building brains, our control over them is limited.
Outcome over process. Ackerman argues creativity should be judged by results. Human art usually carries real emotion, but commissioned professionals deliver regardless, so objections to AI art should be recognised as objections to process.
Trust and hallucination. The oracle image leads people to trust chatbots implicitly; Ackerman says healthy doubt would curb much misinformation and blind advice-taking. She does not treat prediction as a flaw, since humans learn the same way, but says no process yields perfection. Hallucination, in her account, is the shared mechanism of human and machine creativity.
Wave AI's co-creative tools. Lyric Studio runs on a lyrics LLM the company built from scratch and offers suggestions beside the user's own text; Melody Studio proposes chords and melodies. She says people at Universal Music use them. She argues "one hit wonder" GenAI apps fail because users feel unnecessary, and that replacing workers wholesale would collapse the economy.
Models as a mirror. Through Carl Jung's collective consciousness, Ackerman reads models as Western belief made queryable: Midjourney once rendered Hanukkah as a Christmas tree. Her research found image models reproduce brilliance bias, the association of genius with men.
What comes next. She expects AI to become more social and collaborative, with targeted edits such as inpainting rather than full regeneration. The host's failed attempt to get one model to fix a misspelled word in a thumbnail illustrates the gap; she calls these systems "imagination engines" that labs layer controls on top of.
Opening a Thanksgiving week with the market swinging sharply -- big single-week moves in Nvidia, Micron, Oracle, and Tesla, and the NASDAQ up 500-plus points on the day -- the hosts spend the show on a single connecting thread: credentials, vision, and deliberation all matter less than most people assume, while trust, communication, and speed of decision matter more.
Career mailbag. A PhD holder whose degree seems to be hurting the job search is told to consider A/B testing a resume with and without the credential, since hiring managers often assume a PhD candidate is either unaffordable or overqualified; the hosts concede the irony that on their own teams, PhD hires have consistently been "exceptional," even as the broader market frequently penalizes the degree. Asked about the value of getting an MBA immediately after undergrad versus after a few years of work, both hosts favor waiting, citing personal experience (one deferred admission and never returned; the other's son worked several years, found his field didn't fit, then went back) and noting that most MBA cohorts today are dominated by people with prior work experience. A question comparing a semester abroad to an internship draws real disagreement: one says an internship carries far more weight on a resume since a semester abroad "means nothing" to an employer regardless of personal growth, while the other argues travel's maturing effect can outweigh an internship's resume value unless the internship is in a field the student is genuinely committed to. On whether stocks or crypto looks cheaper for a small investor, one host says spreading a small amount across both beats picking one, while the other, already modestly short equities and long crypto after buying Bitcoin near $80,000, leans crypto as the relatively cheaper asset class. On tailoring a resume with AI for a specific job, the advice is to use it but not let it fully replace personal detail, illustrated by an anecdote about a job-seeker whose resume wasn't being read correctly by applicant-tracking software until she switched to a more generic font.
What makes a strong leader. The main segment distinguishes vision from leadership: one host argues that founders and innovators typically hold the vision, while effective leaders are defined instead by the trust they build with a team, pointing out that he and his co-host run very different leadership styles -- one highly transparent, one less so -- yet both say trust from their teams was what came back to them. Flat organizational structures are favored over rigid hierarchies because they preserve the risk-reward incentive for employees to suggest new ideas, though the hosts stress that a flat structure doesn't mean every decision is a group vote. Emotional intelligence -- the ability to "read the room" -- is treated as a related but distinct skill, with both hosts naming Elon Musk and Mark Zuckerberg as examples of leaders they see as lacking it, tying the discussion to a separate conversation about rising autism diagnoses reflecting changed diagnostic criteria rather than a true spike in prevalence. The segment closes on the idea that how someone behaves under pressure, not during success, is what a team actually remembers and what ultimately defines a leader's reputation.
Mega billionaires, and outlier qualities. A scripted rant argues that mega-billionaires risk becoming "material winners and legacy losers," spending their prime years lobbying for tax loopholes rather than building the kind of legacy that costs them something when it counts, and dismissing late-life philanthropic naming gifts as an attempt to buy back reputation rather than genuine impact. The closing segment on outlier qualities -- communicating without ego, demonstrating expertise without seeking attention, staying approachable, differentiating without exaggeration, and deciding quickly -- lands hardest on the last point: the hosts describe closing major business deals in minutes, including one sold before breakfast arrived and another finalized while the other party was in the bathroom, framing fast, confident decision-making, honed through years of trading, as the clearest signal of real potential in a person or a pitch.
Gabriel Dos Reis, principal SDE at Microsoft and a C++ standards committee member since 1997, ends up arguing that modern C++ deserves a fresh look: it evolves by upgrading in place rather than breaking working code, and its remaining gaps, tooling and memory safety by default, are being worked on, though profiles missed C++26.
From fixing GCC to the committee. Dos Reis says he reached the committee after reporting bugs in his 1997 GCC numerics work and being told to join his national body in France. He joined Microsoft in 2013 because the division wanted a standard-conformant compiler, and says no Microsoft dialect of C++ has shipped since.
Why the reputation lags. Because old code still compiles, nothing forced modernization, and advocates never showcased a modern killer app. Hanselman offers an analogy of languages having a temporal dimension, like 1990s Parisian French in Quebec; Dos Reis accepts it, noting MSVC holds 1980s comments beside current features, so his team reverse-engineers old idioms and documents the lessons.
Engineering, not cleverness. He distinguishes programming from engineering, wants code to be "dull" and robust for five to ten years, and teaches problem-first, starting with constexpr by default. Tooling and safety. He concedes the committee neglected package management and build systems; modules in C++20 forced the issue, and he asked Herb Sutter for a tooling study group. On safety he argues the question is scale, not possibility; profiles, developed with Bjarne Stroustrup, and parts of contracts will ship on a parallel track beside C++26.
Math and other languages. He used a mathematical proof to show constexpr was sound when skeptics doubted it, and worked with OCaml's creator on a mechanized proof that RAII is sound. He recommends Stroustrup's A Tour of C++ and learning contrasting paradigms.