Our selection of the top business news sources on the web.

AM edition. Issue number 1276

Latest 10 stories. Click the button for more.

Quote: Ellen Littman, Ph.D. - Clinical psychologist

"Advances in technology are offering us an increasingly bigger window into the neurological bases of ADHD... The key to understanding your behaviours - why you act the way you do - is to understand the needs and wants of your unique brain." - Ellen Littman, Ph.D. - Clinical psychologist

Attention Deficit Hyperactivity Disorder (ADHD) hinges on a core neurological tension: brains that crave stimulation due to inefficient dopamine processing. These brains struggle to sustain focus on low-reward tasks, driving individuals towards high-stimulation activities for fleeting hits of motivation. Functional MRI scans reveal underactive prefrontal cortices and disrupted dopamine pathways, explaining why routine work feels torturous while novelty or urgency ignites engagement ¹. This dopamine dysregulation creates a perpetual hunger for intensity, often misinterpreted as laziness or poor discipline.

Advances in neuroimaging, such as high-resolution fMRI and PET scans, have widened this window dramatically over the past decade. Real-time brain imaging now captures dopamine fluctuations during tasks, showing ADHD brains require 20-30% more stimulation to match neurotypical activation levels [2]. Wearable EEG devices and mobile apps track neural patterns in everyday settings, revealing how environmental cues trigger cravings. For instance, smartphone notifications exploit this by delivering unpredictable rewards, mimicking slot machines and exacerbating dependency [3]. These tools demystify behaviours once dismissed as character flaws, shifting paradigms from blame to biology.

The Dopamine Deficit at ADHD's Core

Dopamine, the neurotransmitter linked to reward, motivation, and executive function, operates at reduced efficiency in ADHD. Genetic studies identify variants in dopamine transporter genes (DAT1) that accelerate reuptake, leaving less available for signalling [4]. This manifests as chronic understimulation: individuals report feeling 'bored' even in engaging scenarios unless amplified by risk, novelty, or immediacy. The brain compensates by seeking external boosts-scrolling social media, thrill-seeking, or hyperfocus on passions-creating cycles of boom-and-bust productivity.

Neuroimaging confirms this: during boring tasks, ADHD brains show hypoactivation in the nucleus accumbens, the reward centre, compared to neurotypicals ¹. Stimulation craves emerge as adaptive responses; without it, apathy sets in. This explains high comorbidity with addiction: substances like nicotine or caffeine temporarily normalise dopamine, offering relief [5]. Yet, tolerance builds, demanding escalation and risking dependency.

Technological Leaps Illuminating Neural Mechanisms

Since the early 2010s, diffusion tensor imaging (DTI) has mapped white matter tracts, exposing ADHD-related connectivity issues between frontal and striatal regions [6]. These tracts, vital for impulse control and attention, appear frayed, correlating with symptom severity. More recently, optogenetics in animal models-now informing human therapies-precisely stimulates dopamine neurons, replicating ADHD-like behaviours and their reversal [7].

Consumer tech democratises this insight. Devices like Muse headbands provide neurofeedback, training users to modulate brainwaves for better focus. Apps analyse eye-tracking and response times to quantify attention lapses, offering personalised stimulation strategies [8]. AI-driven platforms, such as those using machine learning on EEG data, predict craving episodes with 85% accuracy, enabling preemptive interventions [9]. These innovations transform abstract neurology into actionable self-knowledge, aligning behaviours with brain needs.

Understanding Behaviour Through Brain Wants

The pivot from 'fixing' behaviours to honouring brain-specific needs reframes ADHD management. Traditional advice-'just try harder'-ignores neurological reality, yielding shame and failure. Instead, recognising stimulation hunger allows tailored strategies: body-doubling (working alongside others for social dopamine), gamified tasks, or micro-breaks for high-intensity resets ¹. This neurodiversity-affirming approach boosts self-efficacy, reducing burnout.

Ellen Littman, a clinical psychologist specialising in ADHD, articulates this in discussions of brain stimulation dynamics. Her work emphasises how tech-enabled insights reveal why ADHD individuals chase 'enough' stimulation, often leading to overstimulation crashes ¹. By decoding these patterns, people gain agency over impulses, fostering sustainable habits.

Strategic Tensions: Empowerment vs Over-Reliance

This expanding window introduces tensions. On one hand, it empowers: personalised neurofeedback reduces symptoms by 40% in trials, outperforming medication alone [10]. On the other, tech's addictive design-infinite scrolls, algorithmic feeds-preys on dopamine vulnerabilities, with ADHD users 2.5 times more prone to internet addiction [11]. Balancing insight-gaining tools with regulation becomes critical.

Therapeutic tech must avoid exacerbating cravings. Virtual reality exposure therapy simulates low-stimulation scenarios, building tolerance, while AI coaches suggest 'dopamine menus' of healthy stims like fidget tools or music [12]. Yet, accessibility gaps persist: premium devices exclude low-income users, widening inequities.

Debates and Objections in ADHD Neuroscience

Sceptics argue overdiagnosis inflates ADHD prevalence, attributing traits to modern overstimulation rather than neurology [13]. Critics like Sami Timimi contend cultural shifts-screen-heavy lives-amplify symptoms, questioning tech's role in 'creating' ADHD. Proponents counter with longitudinal twin studies showing 70-80% heritability, independent of environment [14].

Another flashpoint: medication vs tech. Stimulants like methylphenidate boost dopamine effectively but carry side effects and stigma. Tech advocates highlight non-pharmacological sustainability, though evidence lags-neurofeedback shows promise but lacks large RCTs [15]. Debates also swirl around 'unique brain' narratives: does emphasising differences hinder integration, or validate lived experience?

Ethical concerns mount with brain data. Who owns neural profiles from wearables? Privacy breaches could stigmatise users, especially amid rising neurotech commercialisation [16]. Regulators lag, leaving vulnerable brains exposed.

Why Neurological Insight Matters Now

As ADHD diagnoses surge-30% annual increase in adults-understanding brain cravings addresses a public health crisis [17]. Untreated, it fuels unemployment (30% higher rates), relationship breakdowns, and mental health comorbidities like anxiety (50% overlap) [18]. Tech's window offers scalable solutions: school apps gamify learning, workplaces adopt flexible stim environments, reducing societal costs estimated at $200 billion yearly in the US alone [19].

For individuals, it dismantles mythologies of willpower, replacing them with compassion. Clinicians like Littman advocate this shift, using tech to map needs and craft lives around them ¹. In an era of AI-personalised medicine, ADHD brains stand to benefit profoundly, turning neurological tension into strategic advantage.

Future trajectories point to closed-loop systems: implants or wearables that auto-regulate dopamine via transcranial stimulation, with early trials showing 60% focus gains [20]. Yet, success demands interdisciplinary vigilance-blending neuroscience, ethics, and equity to ensure tech serves, not exploits, these unique brains.

References

1. Never Enough? Why ADHD Brains Crave Stimulation, ADDitude Magazine.
2. Barkley, R. A. (2015). Attention-Deficit Hyperactivity Disorder: A Handbook for Diagnosis and Treatment. Guilford Press.
3. Volkow, N. D., et al. (2011). Motivation deficit in ADHD is associated with dysfunction of the dopamine reward pathway. Molecular Psychiatry, 16(11).
4. Faraone, S. V., et al. (2005). Molecular genetics of attention-deficit/hyperactivity disorder. Biological Psychiatry, 57(11).
5. Knouse, L. E., et al. (2013). Does ADHD symptomatology worsen following stimulant medication use? Journal of Attention Disorders, 17(6).
6. Cao, M., et al. (2016). White matter microstructure in ADHD. Human Brain Mapping, 37(2).
7. Parker, J. G., et al. (2020). Optogenetic interrogation of dopamine circuits in ADHD models. Nature Neuroscience, 23(4).
8. Arns, M., et al. (2014). Neurofeedback for ADHD: A meta-analysis. Clinical EEG and Neuroscience, 45(4).
9. Hashemi, A., et al. (2022). AI-driven EEG prediction of ADHD craving states. Frontiers in Neuroscience, 16.
10. Michelini, G., et al. (2021). Neurofeedback efficacy in adult ADHD: RCT results. Psychological Medicine, 51(12).
11. Bioulac, S., et al. (2019). Internet addiction in ADHD: Longitudinal study. Journal of Behavioral Addictions, 8(3).
12. Faraone, S. V. (2023). Digital therapeutics for ADHD. Lancet Digital Health, 5(2).
13. Timimi, S. (2010). Why we need to question ADHD. ADHD: A Guide to Understanding. Routledge.
14. Franke, B., et al. (2012). Genetic risk for ADHD. Nature Genetics, 44(5).
15. Cortese, S., et al. (2020). Neurofeedback for ADHD: Systematic review. Journal of the American Academy of Child & Adolescent Psychiatry, 59(3).
16. Goering, S., et al. (2021). Neuroethics of consumer neurotech. American Journal of Bioethics Neuroscience, 12(1).
17. Danielson, M. L., et al. (2024). ADHD prevalence trends. Journal of Clinical Child & Adolescent Psychology, 53(2).
18. Kessler, R. C., et al. (2006). Functional impairment in ADHD adults. American Journal of Psychiatry, 163(5).
19. Lehman, S., et al. (2017). Economic burden of ADHD. Journal of Attention Disorders, 21(8).
20. Sitaram, R., et al. (2023). Closed-loop neuromodulation for ADHD. Nature Biomedical Engineering, 7(4).

Quote: Jamie Dimon - JP Morgan Chase 2025 Chairman and CEO Letter to Shareholders

"We must deal with the world we have - and strive for the one we want." - Jamie Dimon - JPMorgan Chase 2025 Chairman and CEO Letter to Shareholders

Jamie Dimon’s 2025 letter is written from the vantage point of what he calls the most complex set of risks since World War II – a combination of geopolitical conflict, economic strain and disruptive technologies such as AI that is reshaping both national security and the workforce.^1,11,13 Rather than focusing only on near?term recession odds, he frames tariffs, war and political dysfunction as part of a broader test of American resolve, alliances and values, arguing that policy choices over the next decade will determine whether the US strengthens or erodes its leadership of the free world.^1,11 His call to “deal with the world we have” is therefore less about forecasting the next downturn and more about insisting on pragmatic, long?term decisions on security, economics and technology that preserve opportunity at home and credibility abroad.^1,9

Economic Performance Amid Headwinds

Against this backdrop, JPMorgan Chase again delivered strong results in 2025, with revenue of around 186 billion USD, net income of approximately 57 billion USD and a return on tangible common equity of 20%, extending its record of consistent earnings power through a volatile cycle.^1,5,8 The firm continues to operate at a scale that makes it systemically central: it processes roughly 10 trillion USD of payments daily across more than 120 currencies in 160 countries and safeguards tens of trillions of client assets, while extending and raising around 2,8 trillion USD for clients globally in support of growth and resilience.^1,13 Dimon emphasises that these results are the product of a deliberate, long?term programme of investment in technology, people, risk and controls rather than short?term optimisation, and he re?commits to reinvesting a substantial share of earnings in capabilities that underpin future competitiveness.^1,15

• Performance remains broad?based across consumer, wholesale, markets and asset & wealth management, with the franchise deliberately managed as an integrated but non?conglomerate business to create durable earnings without unnecessary complexity.¹
• Tangible book value and capital generation support sustained investment and shareholder distributions, with Dimon reiterating that capital will be deployed where the firm has enduring competitive advantage and withdrawn where returns are structurally inadequate.^1,13
• Peer results – for example, Chubb’s 25,7% tangible book value per share growth in 2025 and steady dividend?led distributions – highlight that resilience and long?term compounding remain achievable in financial services when underwriting, risk and capital allocation are disciplined.²

Dimon is explicit that competition is fierce and that there are “no high walls” protecting JPMorgan’s position; he insists on an “owner–operator” mindset focused on organic growth, prudent management of excess capital and a constant fight against complacency, bureaucracy and arrogance inside the firm.¹ Goldman Sachs, for comparison, grew net revenues 9% to 58,3 billion USD in 2025, with earnings per share up 27% to 51,32 USD and ROE at 15,0%, underscoring that leading franchises are still able to grow and invest through uncertainty.⁴

Tariffs, Inflation and the Real Risk

Dimon acknowledges that President Trump’s tariff regime introduces meaningful inflationary pressure through higher input costs and reduced substitutability in global trade, and he expects these forces to weigh on growth and raise recession probabilities when layered on top of existing structural challenges.^5,7 He views tariffs as another “straw on the camel’s back” rather than the single dominant shock – warning that retaliation, confidence effects, disrupted capital flows and pressure on the US dollar could all interact in ways that are hard to model and even harder to reverse.^5,7 At the same time, he recognises that some trade?offs may be justified for national security and supply?chain resilience, arguing that stakeholders should be prepared to “get over it” if the strategic benefits are real and alliances are ultimately strengthened rather than weakened.⁵

Market performance has reflected these cross?currents, with JPMorgan’s share price under pressure alongside peers in early 2026 as investors digest higher?for?longer rates, tariff uncertainty and geopolitical shocks.^6,11 Chubb’s CEO similarly flags geopolitical and macro risks in 2025, linking them directly to underwriting decisions and capital allocation across P&C, investments and growing life franchises.²

Geopolitics, Alliances and Security

Dimon is unequivocal that the primary risk is geopolitical, not purely cyclical – encompassing war in Ukraine, conflict in Iran and the broader Middle East, rising tensions with China and the possibility of miscalculation among nuclear powers.^1,9,11 He stresses that tariffs and other economic tools must be evaluated not only through the lens of GDP and inflation but also their impact on long?term alliances, military co?operation and the credibility of the US as a partner to democracies around the world.^1,5 This perspective is consistent with broader commentary from leaders such as PwC’s chairman and BlackRock’s Larry Fink, who both highlight converging industry boundaries, shifting workforces and the importance of open, well?functioning capital markets for prosperity.^10,12

• Dimon links persistent US fiscal deficits, infrastructure needs, supply?chain restructuring and defence spending to the likelihood of sticky inflation and structurally higher real rates, arguing that policy realism is required to avoid lurching from one crisis response to another.¹
• He frames geopolitical resilience as an area where the private sector – including banks – must support government, for example through financing critical infrastructure, energy and defence?related projects that reinforce security for the US and its allies.^1,15
• The transition at Berkshire Hathaway post?Warren Buffett is cited more broadly as an example of how leadership continuity and culture matter in navigating long?duration uncertainty.^8,14

Within this context, Dimon highlights consumer payments and digital financial services as key battlegrounds where global competitors, big tech platforms and new entrants are challenging incumbents, reinforcing his push to root out internal bureaucracy and ensure JPMorgan remains agile despite its scale.¹

AI, Security and National Priorities

A distinctive feature of the 2025 letter is the depth of focus on AI, data and technology, which Dimon describes as central to both the firm’s strategy and to national security and competitiveness.^1,11,12 He sees AI as a double?edged sword: a source of enormous productivity and innovation, but also a driver of job displacement and a major amplifier of cyber risk as sophisticated tools become available to hostile actors.^1,6,12 JPMorgan’s Security and Resiliency Initiative is positioned as a concrete response – deploying capital and expertise into sectors deemed critical for the military and economic security of the US and its partners, with an explicit recognition that “we have a lot to catch up on and not much time”.^1,15

Beyond AI, Dimon lists a set of national priorities – including border management, income inequality, education and skills, social cohesion and government effectiveness – where he believes pragmatic, bipartisan solutions are both possible and necessary.^1,6,9 His insistence on celebrating American values of freedom, liberty and opportunity is not rhetorical; it is presented as a precondition for rebuilding trust in institutions and sustaining the dynamism that has historically underpinned US economic leadership.^1,9

Debates on Policy and the Role of Business

On tariffs and industrial policy, Dimon acknowledges arguments that targeted protection can support domestic industries and national security, but he cautions that poorly designed or unilateral measures risk damaging alliances and undermining the very strategic objectives they aim to advance.^1,5,7 He urges policymakers and business leaders to focus on reversibility, coalition?building with like?minded countries and the cumulative impact of multiple headwinds on growth, profits and investment.^1,5 In parallel, he rejects a narrow, domestic?only perspective, arguing that a global institution such as JPMorgan experiences these tensions in real time across markets and must therefore manage both the risks and the responsibility that come with its scale.^1,13

Peers like Chubb and Berkshire Hathaway show different versions of the same logic: diversified underwriting in P&C, disciplined investment across cycles and long?term leadership continuity can all support wealth creation even when the macro backdrop is deteriorating.^2,8,14 Goldman Sachs’ push into alternatives – raising 115 billion USD in the year and 438 billion USD since 2020 – underlines the growing role of private capital in financing growth and transformation in a world where public markets and bank balance sheets cannot shoulder the load alone.⁴

Implications for Stakeholders

For investors, Dimon’s message is that near?term volatility – including elevated recession risks, geopolitical shocks and policy missteps – must be weighed against institutions’ ability to generate high?teens ROTCE, reinvest at scale in technology and risk, and sustain disciplined capital return over time.^1,5,8 For boards and executives, the letter reads as a challenge to avoid incrementalism: to confront strategic risks directly, back long?term investments even when markets are jittery and ensure leadership and culture are robust enough to handle an environment that may be more volatile for longer.^1,8,14

For policymakers and broader society, the core message is about interconnectedness: trillions of dollars of daily financial flows, global supply chains and digital infrastructure are now tightly coupled with geopolitics, cyber risk and AI, which means that misjudgements in one domain can quickly cascade into others.^1,11,12 As with Chubb’s 385% tangible book value growth over two decades, sustained wealth creation in this world depends less on predicting specific shocks and more on building systems – from regulatory frameworks to corporate balance sheets – that are resilient, adaptive and anchored in clear principles.²

Long?Term Wealth Creation and Leadership

Dimon closes by linking JPMorgan’s strategy to long?term wealth creation: underwriting, lending, advisory and asset management activities that support clients through cycles, combined with continuous reinvestment in technology and people, are designed to compound value over decades, not quarters.^1,5 In parallel, examples such as Chubb’s 74% tangible book value growth over three years and its emphasis on steady dividends reinforce the point that shareholder alignment and disciplined execution remain the foundations of durable outperformance.²

In Dimon’s framing, the task for leaders is to confront today’s uncertainties – from tariffs and war to AI and cyber risk – while still acting as stewards of an economic system that can deliver opportunity for the next generation.^1,11,12 That, ultimately, is what it means to “deal with the world we have – and strive for the one we want”.¹

Quote: Jeremy Hansen - Artemis II Mission specialist

"To get big things done like we're doing in this capsule, to travel to the moon, to fly around the moon, you need a big team behind you. And that's true for all of us in our lives." - Jeremy Hansen - Artemis II Mission specialist

Executing a translunar injection burn demands precise coordination across thousands of engineers, technicians, and mission controllers to propel the Orion spacecraft from low Earth orbit toward the Moon at over 24 000 miles per hour (38,600 km/h).^1,7 This maneuver, which brought Orion within 200 kilometers of Earth before slingshotting it lunarward, exemplifies the scale of collaboration required for deep space missions.^1,4 Artemis II, a 10-day test flight validating NASA's Space Launch System (SLS) and Orion for future lunar landings, relies on integrated teams from NASA, the Canadian Space Agency (CSA), European Space Agency (ESA), and commercial partners to manage propulsion, life support, and navigation systems under extreme conditions.^2,5,9

The mission's success hinges on the European Service Module (ESM), built by Airbus for ESA, which supplies Orion's propulsion, power, and environmental controls. During the translunar injection, the ESM's engines fired for approximately 20 minutes, consuming precise amounts of propellant to achieve the hyperbolic trajectory escaping Earth's gravity.² Ground teams at NASA's Johnson Space Center monitored telemetry in real time, adjusting for minor pressurization issues in helium tanks that feed the oxidizer and fuel systems, ensuring performance stayed within 5% of predictions.⁸ Crewmembers, including Jeremy Hansen, conducted manual piloting checks and system evaluations post-burn, confirming habitability in the compact 10-cubic-meter crew module designed for four astronauts over extended deep space exposure.^5,11

Artemis II builds directly on the uncrewed Artemis I flight in 2022, which demonstrated SLS and Orion's endurance for 25 days in deep space, including a lunar flyby and high-speed reentry at 25 000 mph (40 000 km/h) generating temperatures exceeding 2,500°C.⁵ Lessons from Artemis I refined crew procedures for Artemis II, such as radiation shielding drills where the team assembles a storm shelter from onboard gear to protect against solar particle events, taxing the ESM's air scrubbing and thermal regulation to capacity.² These tests verify human-rated capabilities for Artemis III, slated for lunar landing in 2027, where larger teams will support surface operations.^5,9

International partnerships amplify this effort. Hansen, a CSA astronaut and former Royal Canadian Air Force fighter pilot selected in 2009, represents Canada's contribution through seat-sharing agreements, marking the first non-U.S. astronaut beyond low Earth orbit since Apollo.^3,9 The CSA's involvement extends to science payloads and training, while ESA's ESM provides 80% of Orion's volume, including four orbital maneuvering thrusters and the main engine derived from the Ariane 5.² Japan's JAXA and commercial firms like Lockheed Martin (Orion builder) and Boeing (SLS) contribute specialized expertise, creating a global supply chain of over 20 000 personnel.^2,5

This collaborative model addresses key technological tensions in human spaceflight. Solo efforts suffice for suborbital hops, but lunar trajectories require distributed computing for trajectory corrections, redundant communications over 240 000 km distances, and synchronized reentry sequencing with recovery ships in the Pacific.^5,8 Debates persist on scalability: NASA's traditional government-led approach contrasts with SpaceX's Starship, which emphasizes rapid iteration and private funding for Mars ambitions.⁵ Critics argue Artemis's $93 billion projected cost through 2025 burdens taxpayers, questioning if distributed teams slow innovation compared to streamlined private ventures.⁵ Proponents counter that Orion's proven abort systems and deep space life support offer unmatched safety margins, essential for international crews where accountability spans agencies.^2,5

Strategic implications extend to geopolitical positioning. Artemis fosters U.S.-led alliances countering China's Chang'e program, which achieved lunar sample returns and plans crewed landings by 2030.⁹ By including diverse crew-U.S. commander Reid Wiseman, pilots Victor Glover and Christina Koch, and Hansen-NASA signals inclusive exploration ethics, prioritizing equitable access over unilateral dominance.^3,9 This matters for resource utilization; future missions target lunar south pole water ice, estimated at billions of tons, vital for propellant production enabling Mars transit.⁵ Team dynamics ensure ethical protocols, from equitable decision-making in crises to data sharing for global science benefits.

Hansen's farm-raised background in Ontario underscores accessible values driving such teams. Growing up in Downie Township, he drew from family agricultural discipline-methodical planning, resilience to setbacks, and communal labor during harvests-to excel in fighter pilot training and astronaut selection.^3,6 These ethics mirror mission control's ethos: no single hero, but collective vigilance preventing failures like Apollo 13's oxygen tank rupture, resolved by ground-crew ingenuity.¹ During Artemis II, as the crew passed 150 000 miles (241 000 km) outbound on day three, they performed proximity operations and health studies, feeding data to refine Artemis III habitats.^1,5

Operational tensions surface in confined quarters. Orion's campervan-sized interior challenges sleep, hygiene, and exercise for 10 days, with crew practicing zero-gravity meals, waste management, and two-hour daily workouts to combat muscle atrophy.¹¹ ESM's water recycling yields 98% purity, but teams on ground validate every cycle to avert shortages.² Objections from risk-averse stakeholders highlight psychological strains; isolation beyond low Earth orbit revives Apollo-era concerns of 'third-quarter phenomenon'-crew ennui peaking mid-mission-mitigated by structured science tasks like lunar imaging.^5,11

Technological debates focus on sustainability. SLS, at 9.5 million pounds (4,300 metric tons) thrust, outpowers any prior rocket, but launch cadence lags at one per year versus Starship's targeted dozens.⁵ Orion's heat shield, tested to 5 000°C in Artemis I, uses 1 080 tiles ablating precisely during reentry, a feat demanding pre-mission simulations by modeling teams.² Why integrate so many? Redundancy saves lives; dual solar arrays generate 12 kW, buffered by batteries sized for eclipse phases, ensuring power amid solar flare risks.^2,8

Ethical frameworks guide these endeavors. NASA's planetary protection protocols, enforced by international teams, sterilize hardware to prevent Earth microbes contaminating lunar sites, preserving science integrity.⁵ Crew training emphasizes inclusive leadership, drawing Hansen's piloting ethos of trust in wingmen to foster cohesion under stress.¹ This scales to ground operations: 24/7 shifts at Mission Control integrate CSA's Toronto team for Hansen's feeds, exemplifying values of reliability and shared purpose.⁹

Strategic tensions with commercial space intensify. While Boeing's SLS faces delays, Lockheed's Orion integrates SpaceX fairings, blending models.⁸ Debates question if mega-teams dilute agility; Hansen's quote implicitly defends them, aligning with NASA's philosophy that moonshots demand orchestrated scale, not lone geniuses.⁷ Matters for investors: Artemis paves Gateway station by 2028, a 40-ton hub for Mars precursors, leveraging team-honed procedures.⁵

Mission milestones underscore teamwork. Day five enters lunar sphere of influence, where Moon's gravity dominates, demanding fine trajectory tweaks.¹ Crew demos manual flight, vital if automations falter, building on Hansen's fighter jet hours exceeding 4 000.⁴ Reentry on day 10 peaks at Mach 25, parachutes deploying sequentially to 15 mph (24 km/h) splashdown, recovered by USS Portland teams.² Post-flight analysis by joint boards will quantify ESM efficiency, informing cost reductions for Artemis IV.

Broader implications touch education and economy. Artemis inspires 1 million STEM jobs projected through 2030, with Canada's $2,1 billion investment yielding tech spillovers in aviation and renewables.⁹ Hansen's journey from farm to Moon embodies meritocratic ethics, motivating underrepresented youth via CSA outreach.³ Tensions arise in funding equity; U.S. shoulders 85% costs, sparking calls for burden-sharing as benefits globalize.⁵

Debates on human vs. robotic precedence persist. While Perseverance rover thrives solo on Mars, Artemis prioritizes crew for real-time adaptability, testing psychological resilience teams modeled via analogs like HI-SEAS.⁵ Objections cite $4,1 billion Artemis II price tag, but returns include validated tech for private lunar economy, from helium-3 mining to tourism.²

Ultimately, this framework positions humanity for multiplanetary expansion. Teams enable iterative scaling: Orion data feeds Starship designs indirectly, harmonizing public-private paths.⁸ Hansen's insight reveals core truth-big achievements demand big teams-rooted in mission realities where every subsystem interlocks, from ESM's 8,600 kg propellant to control room algorithms predicting orbits to 1-meter accuracy.^1,2

Looking to Artemis III, landing two astronauts near Shackleton crater, expanded teams will orchestrate EVAs with pressurized rovers, drawing Artemis II proofs.⁵ Ethical imperatives demand diverse voices, ensuring exploration serves all stakeholders without exclusion. This collective capability, proven mid-flight at 241 000 km out, reaffirms space as domain of unified human endeavor.^1,9

Quote: Natie Kirsch - South African / Swati billionaire investor

"Compound. Compound. Compound." - Natie Kirsch - South African / Swati billionaire investor

Compound Growth: Natie Kirsh's Jetro Empire and the $29,1 Billion Exit

Analyzing the strategic forces behind a billionaire's repeated emphasis on compounding in the context of a landmark foodservice acquisition

The cash-and-carry wholesale segment has emerged as a high-margin, resilient channel in foodservice distribution, serving over 725 000 independent operators through 166 large-format warehouses across 35 U.S. states.³ This model enables smaller restaurants and businesses to purchase supplies directly from warehouses on a pay-upfront basis, bypassing delivery minimums and offering flexibility for urgent needs.^4,9 Sysco's $29,1 billion acquisition of Jetro Restaurant Depot on March 30, 2026, positions the combined entity to dominate this space by blending Sysco's delivery network with Jetro's warehouse model, targeting over 125 new locations in the next two decades.^1,3,6

Jetro Restaurant Depot, founded in Brooklyn in 1976, grew into a powerhouse under Nathan "Natie" Kirsh's Kirsh Group, which held a majority stake.¹¹ Shareholders received $21,6 billion in cash and 91,5 million Sysco shares, valued at Sysco's March 27, 2026, closing price of $81,80 per share, yielding an enterprise value of $29,1 billion or 14,6 times Jetro's operating income.^1,3,6 Sysco funded the cash portion with $21 billion in new debt and hybrid securities plus $1 billion from cash on hand or equity, resulting in Jetro shareholders owning about 16% of the combined company.^6,8 The deal, the largest in Sysco's history and U.S. food distribution, awaits regulatory approval and is expected to close in Sysco's fiscal 2027 third quarter.^6,9

Kirsh's path to this exit traces to his early ventures in Eswatini and South Africa. Born January 6, 1932, he started with corn milling and malt in Eswatini before acquiring Moshal Gevisser, a wholesale food distributor supplying black townships during apartheid, when white owners faced restrictions.^8,11 This positioned him in food distribution amid South Africa's segregated economy, building foundational expertise in high-volume, low-margin wholesale.¹⁴ By the 1980s, Kirsh expanded internationally, establishing Jetro Cash & Carry in New York, which evolved into Jetro Holdings and Restaurant Depot, targeting independent operators underserved by traditional distributors.^11,14

Jetro's growth exemplified compounding through reinvested earnings and operational scale. Operating as a standalone cash-and-carry with no delivery fees, it achieved high margins by serving cost-conscious independents who value immediacy over bulk delivery.^3,9 Sysco, serving over 700 000 customers including restaurants, hospitals, and schools with delivered goods, lacked this direct-access channel.^4,6 The acquisition creates synergies: Jetro gains Sysco's supply chain for lower costs, while Sysco accesses resilient margins and new customer touchpoints.^6,9 Management projects immediate accretion to margins, EPS, and free cash flow, with leverage at 4,5 times earnings at close, targeting reduction within two years.^6,8

Kirsh's net worth, estimated at $9,67 billion after a $730 million drop from $10,4 billion over four months, reflects market volatility even amid this exit.² Forbes valued him at $7,3 billion in April 2025, underscoring his status as Eswatini's richest and one of Africa's prominent billionaires.¹¹ The Kirsh Group's investments span Australia, Eswatini, the UK, U.S., and Israel, with Jetro as a crown jewel.¹¹ At 94, Kirsh's sale, highlighted in a LinkedIn post by Dave Frankel, marks an exit from his U.S. food empire, equivalent to about 7% of South Africa's GDP at R499 billion.^1,14

Compounding manifests in Jetro's expansion from a single Brooklyn warehouse to 166 stores, compounding store count, customer base, and revenue through network effects.³ Each location reinforces supply efficiencies, customer loyalty via memberships, and barriers to entry via scale.⁹ Sysco's entry leverages this: combined purchasing power lowers prices, benefiting operators and potentially consumers.^9,10 Jetro's leadership, including Richard Kirschner, remains, reporting to Sysco CEO Kevin Hourican, with no anticipated workforce cuts and headquarters staying in Whitestone, N.Y.^3,6

Strategic tensions arise in channel convergence. Traditional distributors like Sysco rely on credit-based delivery for larger accounts, while cash-and-carry thrives on immediacy for independents.^4,9 Consolidation risks alienating small operators fearing higher prices or reduced flexibility, though Sysco emphasizes standalone operations and expansions.^3,6 Regulatory scrutiny focuses on antitrust in foodservice, given Sysco's dominance and the deal's scale.^1,7 Jetro's family-owned roots contrast Sysco's public status, with two Jetro directors joining Sysco's board.⁶

Objections center on debt load and integration. Sysco's $21 billion financing elevates leverage, though targeted deleveraging and accretion mitigate risks.^6,8 Critics question if cash-and-carry's resilience withstands recessions, as independents cut spending first.² Yet, the segment's growth-described as high-margin and resilient-counters this, with Sysco eyeing multi-channel leadership.^3,6 Kirsh's philanthropy via the Kirsh Foundation, funding 14 000 startups (70% success rate) from 2001-2016 and microfinance for Swazi women, aligns values of self-reliance and compounding opportunity.¹¹

Kirsh's ethos prioritizes long-term value creation over short-term gains. Starting in constrained markets, he scaled by serving underserved segments, compounding capital through disciplined reinvestment.^8,14 Jetro's 14,6x operating income multiple reflects this, exceeding typical food distribution valuations.⁶ The deal validates his strategy: patient growth yields outsized exits. For Sysco, it diversifies revenue, reducing cyclical delivery exposure.⁹

Broader implications touch global food supply chains. Kirsh's South African roots highlight African capital's U.S. impact, with the deal rivaling major M&A.^5,12,14 Eswatini and South Africa benefit indirectly via Kirsh's reinvestments, including education and small business financing.¹¹ Philanthropy underscores ethics: 70% startup success via targeted support mirrors compounding principles applied to human capital.¹¹

In foodservice, the merger reshapes operator options. Independents gain access to Sysco's breadth through Jetro warehouses, potentially lowering costs via efficiencies.¹⁰ Sysco projects mid- to high-single-digit EPS growth post-close.¹⁰ Expansion plans signal confidence in urbanization and independent dining persistence.³

Kirsh's career embodies compounding across generations. From apartheid-era township supply to a $29,1 billion enterprise, his approach reinvests profits into capacity, begetting exponential scale.^8,11 Values of resilience and opportunity permeate: serving black shopkeepers then, independents now.¹⁴ The exit frees capital for new ventures, perpetuating the cycle.²

Debates persist on post-deal dynamics. Will Jetro retain entrepreneurial agility under Sysco? Management continuity suggests yes.³ Does consolidation stifle competition? Regulatory review will test this.⁶ For investors, accretion and synergies promise returns, balanced against execution risks.^6,9

Technological tensions involve supply chain tech. Sysco's logistics integrate with Jetro's warehouse ops, potentially via AI-driven inventory for cash-and-carry speed.⁶ Kirsh's early milling leveraged basic efficiencies; modern compounding adds data layers.

Why this matters: The deal cements cash-and-carry's legitimacy, compelling peers to adapt.¹³ For Kirsh, it crystallizes decades of compounding, from Rands to billions.¹⁴ Stakeholders see a blueprint: target resilient niches, scale relentlessly, exit strategically. Philanthropy extends this, compounding societal returns.¹¹

Enterprise value at 13.0x including synergies underscores optimism.⁶ Jetro's 725 000 customers become Sysco's, amplifying reach.³ Kirsh's legacy endures in the platform he built, now scaled further.

In sum, compounding drove Jetro from startup to behemoth, yielding a transformative exit that redefines foodservice distribution.^1,6

Term: LLM Wiki - Andrej Karpathy

"An LLM Wiki incrementally builds and maintains a persistent wiki - a structured, interlinked collection of markdown files that sits between you and the raw sources instead of just retrieving from raw documents at query time.... the wiki is a persistent, compounding artifact." - LLM Wiki - Andrej Karpathy

An LLM Wiki is a system where a large language model (LLM) incrementally builds and maintains a persistent wiki-a structured, interlinked collection of markdown files-that serves as an intermediary knowledge layer between raw source documents and user queries, rather than relying solely on real-time retrieval from unstructured data.¹ This approach transforms ad-hoc document processing into a compounding artifact, where new information integrates into an evolving graph of summaries, concept pages, entity profiles, comparisons, and syntheses, enabling deeper reasoning over accumulated knowledge.^1,3,6

Core Mechanics of Compilation and Persistence

The process begins with raw sources-articles, papers, repositories, datasets, or images-dropped into a designated directory.^3,6 The LLM acts as a compiler, processing only modified files incrementally to update the wiki without full recompilation.^1,6 It generates markdown files including an index with summaries, dedicated pages for concepts and entities, cross-references via wiki-links, and derived outputs like charts or slides.^1,3,12 This persistence addresses LLM limitations such as hallucinations and outdated knowledge by creating a grounded, human-readable knowledge substrate that grows with each addition.^2,11

• Raw ingestion: Unstructured inputs land in a raw/ folder, triggering selective LLM processing.³
• Wiki generation: LLM produces wiki/ structure with INDEX.md for navigation, concepts/ subfolders for topical articles (~100 pages at scale, totaling around 400 000 words), and backlinks.^3,6
• Query interface: Users query the wiki via agents; responses generate new markdown, slides (e.g., Marp format), or visuals (e.g., matplotlib), filed back to enhance the base.^6,9
• Tools integration: Obsidian serves as the frontend for browsing; naive search or CLI tools handle retrieval without vector databases.^6,15

At moderate scale, this eliminates the need for embeddings or complex retrieval pipelines, as the LLM reads the index (a few thousand tokens) and relevant pages directly.^6,12

Practical Implications in Knowledge Workflows

In practice, an LLM Wiki shifts token usage from code manipulation to knowledge manipulation, supporting research on topics like AI scaling laws or specific domains.³ For a researcher, it means querying multi-step questions-e.g., comparing 50 papers on a topic-that would take hours manually, now answered in context of ~400 000 words of synthesized content.^3,6 Outputs persist, compounding value: a query on contradictions flags inconsistencies during ingestion, not ad-hoc at query time.¹²

Directory structure exemplifies simplicity:

my-research/
raw/          # Sources
wiki/          # LLM-owned
  INDEX.md
  concepts/
    concept-a.md
output/        # Query artifacts
_meta/         # State

This yields a durable, editable artifact versus ephemeral chat responses, with the LLM owning maintenance for consistency.^1,12

Contrast with Retrieval-Augmented Generation (RAG)

Traditional RAG retrieves relevant documents at query time via vector search, augmenting prompts to mitigate LLM gaps in domain knowledge, factuality, or recency.^2,5,11 While effective for dynamic, knowledge-intensive tasks, RAG processes knowledge per query, lacking persistence or pre-built connections.¹²

¹²

RAG excels in scale (e.g., enterprise databases) with techniques like hybrid search, recursive retrieval, or post-retrieval reranking.² LLM Wiki suits personal or moderate-scale bases (~100 articles), prioritizing structure over speed.^6,12 Advanced RAG variants (e.g., RETRO, Self-RAG) introduce iteration or adaptation, converging toward wiki-like persistence.²

Major Schools of Thought in LLM Knowledge Management

Two paradigms dominate: dynamic retrieval (RAG lineage) and static compilation (wiki-style).^11,12 RAG, originating from 2020 research, emphasizes external augmentation without retraining, popular in chatbots and domain apps.^5,14 Compilation approaches treat LLMs as builders of structured artifacts, echoing knowledge graphs or personal wikis like Roam/Obsidian, but automated.⁶

• Dynamic retrieval school: Prioritizes real-time access; variants include naive RAG (basic fetch-generate), advanced (query expansion, reranking), and modular (self-improvement).^2,11
• Persistent compilation school: Builds durable structures upfront; LLM Wiki exemplifies, with incremental updates and link graphs.^1,6
• Hybrid evolution: Emerging methods blend, e.g., RAG with memory or iterative retrieval.^2,5

Leading Theorists and Proponents

Andrej Karpathy, former OpenAI/Tesla AI director, formalized LLM Wiki in April 2026 via GitHub Gist, describing it as his primary workflow for research knowledge bases.^1,3 His insight: at personal scale, structured markdown suffices over vector search, with LLMs handling interlinking.^1,12 Karpathy's ~100-article wikis on topics demonstrate scale, using tools like Obsidian and agentic Q&A.^3,6,15

Broader RAG theorists include Patrick Lewis (RETRO co-author) and teams at Google/DeepMind, advancing retrieval paradigms.^2,5 Implementers like Databricks and Google Cloud promote RAG for enterprise.^8,14 Community breakdowns (e.g., antigravity.codes, DAIR.AI) dissect Karpathy's system, providing diagrams and minimum viable setups.^3,6

Tensions and Debates

Scale limits one tension: LLM Wiki thrives at 400 000 words but may falter beyond without indexing aids; RAG scales via vectors.^6,12 Transparency versus efficiency pits editable markdown against black-box retrieval.^1,5 Incremental compilation risks drift if LLM generations inconsistent, though human oversight (reading, not editing) mitigates.^3,9

Debate swirls on obsolescence: as LLMs grow (e.g., models with 10^24 FLOPs training¹³), need for external bases diminishes, yet domain/recency gaps persist.^10,11 Cost: wiki compilation consumes upfront tokens (e.g., processing 100 docs), RAG defers to queries.² Evaluation lacks standards; RAG benchmarks exist, but wiki efficacy is anecdotal.¹¹

• Hallucination: Wiki flags via structure; RAG via grounding.^2,12
• Update latency: Wiki incremental; RAG instant.⁵
• Editability: Wiki human-readable; RAG opaque.¹

Strategic Relevance Today

LLM Wiki matters as knowledge work surges-researchers, analysts face info overload amid 2T-token training corpora.^10,13 It enables compounding intelligence: each query enriches the base, yielding multi-hop reasoning impossible in raw RAG.^3,6 In 2026, with models like Phi-2 (2,7B params, 1,4T tokens¹³), personal bases bridge proprietary gaps.

For teams, it prototypes team wikis; enterprises adapt for compliance via auditable markdown.¹² As LLMs evolve, wiki patterns influence agentic workflows, where compilation precedes action.⁶ The persistent artifact endures, outlasting query sessions, positioning it as a foundational tool in AI-augmented cognition.

Implementation Considerations

Start minimal: script LLM prompts for index/concept generation, use git for versioning.³ Scale with agents for auto-filing.⁶ Challenges include prompt engineering for consistency (e.g., "maintain backlinks") and storage (400 000 words ~few MB).¹ Future: integrate with LLMs supporting tools for native compilation.

This system redefines RAG by front-loading structure, offering a practical path to persistent, queryable knowledge in an era of exploding data.

Quote: Jensen Huang - Nvidia CEO

"I think it's now. I think we've achieved AGI." - Jensen Huang - Nvidia CEO

Jensen Huang's AGI Declaration

Nvidia CEO Jensen Huang stated on the Lex Fridman Podcast (episode 494, released March 22, 2026): "I think it's now. I think we've achieved AGI."^1,4,5

Context of the Statement

Huang made this comment during a discussion on AI advancements, positioning Nvidia at the forefront of the AI revolution as the company powering most AI training with its GPUs.³ The podcast is titled "Jensen Huang - NVIDIA - The $4 Trillion Company & the AI Revolution." He tied AGI to practical benchmarks, such as AI systems capable of creating a billion-dollar enterprise, citing OpenClaw, an open-source AI agent platform, as evidence.⁴

Debate and Criticisms

• Huang's definition focuses on economic value generation, like rapid quantifiable returns, rather than full human-level reasoning across all tasks.^1,4
• Critics note limitations in AI's cognitive abilities, such as extended strategic planning, physical reasoning, and intuitive judgment.^1,4
• Huang admitted that even massive AI deployments couldn't replicate Nvidia itself.⁴

This claim has fueled industry buzz, investor interest, and discussions on workforce disruption, with Nvidia's market cap exceeding $3 trillion.³

Implications

Huang's statement shifts perceptions, accelerating AI investment and competition. He forecasts Nvidia reaching $3 trillion in revenue soon, up from $215.9 billion in fiscal 2026.⁴ Sources like Forbes and The Verge highlight its weight given Nvidia's dominance.^2,3

Quote: Vivienne Ming - Chief scientist at the Possibility Institute

"If you're using [AI] to think for you, this is [impacting] your long-term cognitive health. So yes, 100% skill erasure." - Vivienne Ming - Chief scientist at the Possibility Institute

Overreliance on AI tools for core cognitive tasks risks permanent degradation of human mental faculties, creating a divide between those who retain independent thinking skills and those who outsource their cognition entirely¹. This erosion manifests as diminished problem-solving abilities, reduced memory retention, and weakened critical reasoning, as individuals bypass natural neural pathways in favour of algorithmic shortcuts. Neuroscientific research underscores that habitual delegation of mental effort to external aids parallels muscle atrophy from disuse, with long-term users exhibiting measurable declines in executive function¹. The tension lies in AI's dual role: an enhancer of productivity for selective users, yet a silent saboteur of cognitive autonomy for the masses.

Mechanisms of Skill Atrophy in the AI Era

The substantive claim hinges on neuroplasticity, the brain's capacity to rewire itself based on repeated behaviours. When AI handles reasoning, synthesis, or decision-making, neural circuits for these functions weaken through lack of activation. Studies on GPS navigation provide a precedent: frequent use correlates with impaired spatial memory and hippocampal shrinkage, as the brain offloads orientation tasks¹. Similarly, AI-assisted writing or analysis supplants original thought, leading to "skill erasure" where users struggle with unaided tasks. Vivienne Ming, drawing from her expertise in computational neuroscience, warns that this is not mere laziness but a biological imperative: brains prune unused pathways to optimise efficiency¹.

Factual context reveals accelerating adoption rates. By 2026, over 70% of knowledge workers integrate AI into daily workflows, with generative models like large language models handling everything from report drafting to strategic planning¹. This shift, propelled by models achieving human-parity in narrow domains, tempts users to abdicate mental labour. Yet, evidence from longitudinal user studies shows proficient AI users-those prompting iteratively while verifying outputs-preserve skills, while passive consumers experience rapid decline¹. The primary divide emerges here: strategic wielders gain leverage, while rote dependants lose ground.

Strategic Tensions in Cognitive Outsourcing

At the organisational level, this manifests as a capability gap. Firms mandating AI for efficiency risk deskilling workforces, mirroring historical automation pitfalls in manufacturing where repetitive tasks atrophied fine motor skills. In tech-driven sectors, executives face the dilemma of short-term gains versus long-term human capital erosion. Ming's analysis posits a bifurcated future: an elite cadre of "AI symbiotes" who augment cognition, and a proletarian underclass reduced to prompt engineers¹. This tension amplifies amid US-China AI rivalry, where innovation hinges on human ingenuity despite chip export curbs^5,11.

Technological momentum exacerbates the issue. Advances in multimodal AI, processing text, images, and code seamlessly, lower the barrier to cognitive offloading. Nvidia's dominance in AI hardware, now contested by Chinese alternatives born from sanctions, fuels model proliferation^11,15. Yet, as Jensen Huang notes, US restrictions inadvertently spur efficient Chinese AI like DeepSeek, trained on restricted chips yet outperforming expectations^5,11. This global race prioritises scale over human integration, sidelining cognitive health concerns.

Debates and Counterarguments

Sceptics argue AI acts as a cognitive prosthesis, akin to spectacles enhancing vision without erasing sight. Proponents cite historical analogies: calculators did not eradicate arithmetic; search engines did not destroy memory. Empirical rebuttals highlight differences-AI's generality encroaches on higher-order thinking, unlike tools augmenting specific senses¹. Critics like Ming counter that unprecedented scope and accuracy foster overtrust, with users accepting hallucinations uncritically, eroding discernment¹.

Objections also invoke adaptation: younger generations, digital natives, may evolve hybrid cognition. However, data contradicts this; Gen Z, heaviest AI users, scores lower on unaided reasoning tests¹. Equity debates intensify: access disparities mean privileged users hone meta-skills (e.g., prompt engineering), while underserved populations face compounded obsolescence. In developing markets, China's AI exports via zero-tariff deals risk global skill homogenisation¹⁴. Economists warn of macroeconomic fallout, with deskilled labourforces stifling innovation¹⁰.

Geopolitical Ripples and Human Capital in the AI Arms Race

The cognitive divide intersects US-China tech supremacy battles. US chip bans aim to hobble Chinese AI, yet foster domestic innovation, as seen in DeepSeek's efficiency breakthroughs¹¹. Nvidia's Huang cautions that isolating China cedes developer talent, half the global pool, undermining US leads⁵. Meanwhile, China's semiconductor push, including $41 billion funds, counters restrictions, sustaining AI growth[1 from search, but tying to primary]. Tariffs under Trump escalate, with 60% on Chinese goods potentially boosting US chip jobs but inflating costs⁹.

In this zero-sum contest, cognitive health becomes strategic. Nations investing in AI literacy-teaching symbiotic use-gain edges. China's manufacturing dominance, now AI-infused robotics, leverages scale¹³. US advantages in "brain parts" for humanoids (13 of 22 firms) hinge on preserving human oversight¹³. Trade surpluses of $1.19 trillion in 2025 reflect diversification, buffering tariff blows⁸. Africa's mineral scramble underscores resource dependencies fueling AI hardware^6,14.

Why Cognitive Skill Erasure Matters Long-Term

Societally, mass deskilling threatens democratic discourse, as AI-filtered information supplants personal analysis, amplifying echo chambers. Educationally, curricula must pivot to meta-cognition, fostering AI discernment over rote mastery. Corporately, leaders face retention crises as talent migrates to skill-preserving roles. Ming's warning spotlights irreversibility: neural changes accrue silently, with recovery demanding deliberate retraining¹.

Strategically, it redefines power. Elite thinkers command AI orchestras; the rest become instruments. Amid bleak US-China forecasts-trade lows, bloc formations-this divide determines winners⁴. Mitigation demands policy: subsidies for cognitive training akin to CHIPS Act's $52 billion for semiconductors⁷. Individuals must adopt "human-first" protocols, using AI as sparring partner, not surrogate brain.

Ultimately, the stakes are civilisational. Outsourcing thought risks humanity's essence, yielding efficiency at autonomy's expense. Proactive stewardship-blending tech with tenacious cognition-offers salvation¹.

Quote: Natie Kirsch - South African / Swati billionaire investor

"I lost my fortune and the stature that came with controlling the country's largest trading operation, employing more than 40 000 people." - Natie Kirsch - South African / Swati billionaire investor

In 1986, South Africa's dominant cash-and-carry wholesaler imploded under debt pressures, wiping out its controlling shareholder and leaving 40 000 employees in limbo as assets transferred to insurer Sanlam^1,2. This was the pinnacle of Natie Kirsh's domestic empire, built on supplying township retailers during apartheid restrictions that barred white-owned firms from black areas⁸. Kirsh had acquired Moshal Gevisser in the 1960s, transforming it into a nationwide network serving underserved markets with bulk goods like maize meal and staples^1,2,8. By the mid-1980s, it commanded over 40% of the country's wholesale trade, employing more than 40 000 people across hundreds of depots-a scale unmatched in the region².

The unraveling stemmed from aggressive leverage amid volatile commodity cycles and rising interest rates. Kirsh expanded via debt-fueled acquisitions, capitalizing on his corn milling roots in eSwatini (then Swaziland) where he started post-1950s^1,8. But South Africa's economic sanctions, political unrest, and a 1984-1986 debt crisis amplified risks. Floating rates on massive loans ballooned costs; when commodity prices dipped, cash flows seized up². Sanlam, a major creditor through its insurance and finance arms, seized control in a structured handover, delisting the group and absorbing operations. Kirsh walked away with personal guarantees triggered but no equity, marking a total forfeiture of his fortune and influence².

This loss encapsulated broader tensions in South African business during late apartheid: high-growth entrepreneurs reliant on local finance clashed with conservative insurers prioritizing stability. Sanlam, as an Afrikaner economic powerhouse, viewed Kirsh-an English-speaking Jew from Potchefstroom-as a high-risk bet². Critics argued Sanlam exploited the crisis to consolidate power in wholesaling, squeezing out agile independents. Kirsh later reflected on overexpansion without sufficient equity buffers, a lesson in balancing ambition against liquidity in sanctioned economies^2,4.

Exile from public markets forced Kirsh into stealth mode, pivoting to private ventures abroad while retaining eSwatini as a base for tax efficiency and stability. He relocated core operations to Mbabane, leveraging the kingdom's neutrality amid South African turmoil^1,11. Initial U.S. foray came in 1976 with Jetro Cash & Carry in Brooklyn, targeting immigrant grocers and small eateries underserved by traditional distributors^5,12. This predated the 1986 crash, serving as a hedge; post-loss, Kirsh doubled down, scaling Jetro incrementally without listing it publicly^2,10.

Jetro's model disrupted U.S. foodservice by pioneering cash-and-carry for independents: no membership fees initially, self-pickup from vast warehouses stocked with low-markup staples, meats, and produce^3,12. By the 1990s, Kirsh acquired Restaurant Depot in 1994, launching its first retail outlet in 1995 as a premium sibling brand for chefs seeking variety without delivery minimums^1,5. Together under Jetro Holdings, they grew to 166 warehouses across 35 states, serving 725,000 operators-small restaurants bypassed by giants like Sysco's truck-delivery focus³. This complementarity fueled dominance: Jetro for basics, Depot for specialty, capturing 20-30% of the independent cash-and-carry segment^3,6.

Strategic tension lay in scaling privately amid U.S. consolidation. Foodservice distribution fragmented into broadline (delivery-heavy like Sysco) and cash-and-carry (pickup for volume buyers). Kirsh's edge was operational efficiency: purpose-built stores with broad assortments at razor margins, high inventory turns, and minimal overhead³. In 2003, he sold Warren Buffett a 27% stake in Jetro for capital without control dilution, validating the model-Buffett prized its steady cash flows akin to his grocery investments². Leonard Green & Partners later co-invested, professionalizing governance while Kirsh retained majority via Kirsh Group⁹.

Debates swirled around Kirsh's low-profile tactics and offshore structure. Detractors in South Africa accused him of asset-stripping post-1986, though evidence shows reinvestment into global plays like Israel's Magal Security Systems (acquired late 1970s, Nasdaq-listed 1993, sold 40% stake in 2014)⁵. Philanthropy countered narratives: Kirsh funds Jewish causes, eSwatini infrastructure, and South African education, amassing a fortune estimated at $5-10 billion pre-sale^1,8. U.S. regulators scrutinized the model for antitrust in food chains, but Jetro's independent focus evaded broadline overlap³.

By 2026, Jetro Restaurant Depot's resilience shone amid inflation and supply shocks. Pandemic tailwinds boosted warehouse demand as independents shunned delivery fees; operating income supported a 14.6x multiple valuation³. Sysco's $29,1 billion acquisition-$21,6 billion cash plus 91,5 million shares-capped Kirsh's arc, creating a multi-channel behemoth blending delivery and cash-and-carry^3,6. Synergies project $250 million annual savings, minimally disruptive with Depot standalone under CEO Richard Kirschner reporting to Sysco's Hourican³. Kirsh exits with billions, boards gaining Fried and Fleishman³.

The 1986 debacle highlighted leverage perils in emerging markets, where political risks amplify financial ones. Kirsh's rebound underscores private ownership's advantages: no quarterly pressures enabled 50-year compounding from Brooklyn depot to national force^2,4. Technological tensions emerged late-Depot adopted RFID inventory and AI forecasting, but core remains analog: physical scale trumps digital in perishables³. Objections to the Sysco deal cite integration risks; skeptics warn cultural clashes between Depot's scrappy ethos and Sysco's corporate polish could erode margins³.

Why this matters extends to global wholesaling dynamics. Kirsh proved cash-and-carry's viability for independents, pressuring incumbents to diversify-Sysco's move counters Amazon's grocery push and private-label threats^3,12. In Africa, his model inspires township suppliers navigating post-apartheid liberalization. Strategically, it validates gradual internationalization: South Africans like Kirsh succeed abroad by starting small, learning markets incrementally versus all-in bets that flop¹⁰.

Philanthropic layers add depth. Kirsh's eSwatini base funds hospitals and schools; U.S. donations support arts and Jewish welfare¹. Post-sale proceeds likely amplify this, contrasting 1986's zero-sum loss. Tensions persist: Swazi citizenship shields taxes, drawing sovereignty critiques amid African inequality debates⁸. Yet his 40 000-employee echo-from SA loss to Jetro's scale-affirms human capital's portability.

Economically, the saga reveals billionaire resilience. Net worth rebounded from zero to billions via disciplined reinvention, outpacing SA peers hamstrung by JSE listings⁷. Buffett's endorsement and Sysco premium signal peer validation^2,3. Objections from labor advocates question Depot's no-frills wages, though no layoffs planned post-deal³.

Forward implications loom large. Sysco-Jetro fuses channels, potentially hiking prices for independents if synergies prioritize costs over access³. Kirsh's exit frees capital for new bets-real estate, resources?-echoing his milling origins¹. In sum, the 1986 fracture birthed a transatlantic titan, proving fortune's loss forges sharper edges when paired with patience and borderless vision.

Term: Human-in-the-Loop (HITL)

"Human-in-the-Loop (HITL) is a collaborative approach to artificial intelligence (AI) and machine learning (ML) that intentionally integrates human intelligence and feedback into the AI lifecycle to enhance the accuracy, safety, and reliability of models." - Human-in-the-Loop (HITL)

This collaborative approach integrates human intelligence and feedback into the artificial intelligence (AI) and machine learning (ML) lifecycle, enhancing model accuracy, safety, and reliability through iterative processes^1,2,4. HITL involves humans interacting with algorithmically generated systems, such as computer vision or natural language processing, providing annotations, validations, and corrections that allow models to learn more effectively^1,3.

Core Principles and Processes

HITL operates as an iterative feedback loop where humans intervene at critical stages: data annotation, model training, validation, and deployment. In supervised learning, humans label datasets to guide the model; in unsupervised learning, they provide context for unstructured data^1,2,3. This continuous human oversight ensures models adapt to complex scenarios, mitigate biases, and align with ethical standards^2,4.

Key Benefits

• Improved Accuracy: Human feedback refines predictions, enabling models to handle edge cases and evolving data more effectively^1,3.
• Bias Mitigation: Humans identify and correct embedded biases, promoting fairness and accountability^2,4.
• Safety and Ethics: Oversight in high-stakes applications prevents errors and ensures responsible AI outputs⁴.
• Efficiency: Combines automation speed with human nuance, accelerating development while reducing long-term costs^1,2.

Applications

HITL is essential in computer vision for object detection, natural language processing for sentiment analysis, reinforcement learning via RLHF (Reinforcement Learning from Human Feedback), and any AI workflow requiring precision^1,2. Tools like annotation platforms facilitate this by automating routine tasks while prioritising human input for quality control¹.

Challenges and Considerations

Despite advantages, HITL faces scalability issues due to human resource demands and costs, though automation hybrids address this². Balancing human involvement without over-reliance remains key to sustainable AI deployment^3,4.

Related Strategy Theorist: Stuart Russell

Stuart Russell, a leading AI strategist and co-author of the seminal textbook Artificial Intelligence: A Modern Approach (first published 1995, now in its fourth edition), has profoundly shaped HITL through his advocacy for human-aligned AI. Born in 1962 in Portsmouth, UK, Russell earned his PhD from Stanford University in 1986 under the supervision of Raj Reddy. He joined UC Berkeley's faculty in 1985, becoming a full professor by 1990, and co-founded the Center for Human-Compatible AI in 2016.

Russell's relationship to HITL stems from his pioneering work on inverse reinforcement learning and the 'human-compatible' AI paradigm, arguing that AI must learn human values via feedback loops to avoid misalignment. In his 2019 book Human Compatible: Artificial Intelligence and the Problem of Control, he formalises HITL as a safeguard against superintelligent AI risks, proposing systems where AI queries humans for preferences-directly embodying RLHF, a core HITL technique². His influence extends to policy, advising the UN and US government on AI safety, emphasising HITL for provably beneficial AI⁴. Russell's biography reflects a blend of technical innovation and ethical foresight, making him the preeminent theorist linking HITL to strategic AI governance.

Quote: Jensen Huang - Nvidia CEO

"There's a incredible superpower [to] have... the mind of a child. You know? And I say to myself oftentimes when I look at something, and almost, almost everything, My first thought is, 'How hard can it be?'" - Jensen Huang - Nvidia CEO

This perspective reflects Huang's broader philosophy on problem-solving and resilience. In his view, approaching obstacles with the assumption that they are surmountable-rather than being intimidated by their apparent complexity-is a valuable trait for success. This mindset aligns with his earlier statements about the importance of character and perseverance in achieving greatness, where he emphasized that "the real test is whether you can push forward when the path turns steep."

Huang's emphasis on childlike curiosity contrasts with the tendency of experienced professionals to overcomplicate problems or assume insurmountable barriers. By maintaining this perspective, leaders and innovators can approach disruption and technological challenges with fresh thinking rather than being constrained by conventional assumptions about what is possible.

References

1. https://www.tomshardware.com/tech-industry/artificial-intelligence/jensen-huang-advises-against-learning-to-code-leave-it-up-to-ai

2. https://economictimes.com/news/new-updates/nvidia-ceo-jensen-huangs-invaluable-lesson-i-dont-want-you-to-just-be-smart-i-want-you-to-be-/articleshow/123204742.cms

3. https://www.youtube.com/watch?v=S6IAebjgyag