The Didi & Lital Show

In this Lital-less episode of the DiDi and Lital Show, DiDi interviews Denis Knjazihhin, a cybersecurity veteran who founded D2 to help enterprises roll out AI agents about his new company, D2, after a brief sports and music discussion including World Cup reactions and concerts at Fenway. Denis explains that D2 helps organizations adopt AI by removing security as a blocker, since agents require deep access to systems and data and can make risky mistakes or be tricked into harmful actions. D2’s approach is intent-based policy: expressing what should happen and translating it into controls that constrain agent actions deterministically, evolving as agents find new ways around restrictions. A key use case is governing Claude CoWork access to sensitive email in a financial institution with policies like limiting access to recent emails, excluding attorney-client privilege content, restricting to specific folders, and allowing drafting but not sending. D2 deploys via a gateway between the agent and applications, supports local/hosted/enterprise installs, offers natural-language policy creation with testing, and provides an open-source policy catalog. Denis invites users to try D2 free at d2.ai, shares that the team is six people, and says they are seeking GTM leadership and critical product feedback.
 
Topics
01:35 World Cup Takes
03:53 US Team Rollercoaster
07:21 Red Sox and Fenway Concerts
09:39 What D2 Does
14:22 Claude Email Policy Use Case
17:00 Gateway Deployment and Policy Catalog
19:29 Hardening Policies and User Partnership
22:51 Try D2 and Core IP
25:01 PDP vs PEP Architecture
26:49 Small Team Big Speed
27:56 Startup Building Then vs Now
31:45 Wrap Up and Call to Action

Didi hosts without Lital and welcomes returning guest Denis Knjazihhin, a cybersecurity veteran who founded D2 to help enterprises roll out AI agents. Denis recounts being born in the Soviet Union (now Estonia), moving through Germany to Boston, and his startup path from BladeLogic to Aveksa (acquired by RSA), where he met DiDi and later partnered with him on Cisco’s internal startup CDO. They discuss how product success at Cisco depends heavily on integrating with Cisco’s go-to-market machine, highlighting an ASA upgrade feature that drove adoption and services. The conversation shifts to software process: their preference for one-week sprints for rapid feedback, and how AI is disrupting agile practices, moving work from code completion to agent-driven design and implementation guided by ADRs, with growing need for higher-level review and summaries. Denis cites a 153k-line agent-assisted PR done largely during vacation, while both caution against letting AI change core, high-risk systems.
 
Topics
00:52 Denis Returns
02:51 Denis Origin Story
03:55 Pets And Parenting
05:07 Startup Journey Together
07:50 Patents And AI
10:45 Cisco GTM Lessons
12:26 ASA Upgrade Breakthrough
15:24 Startup Strategy Analogies
18:10 One Week Sprint Rule
20:31 Fast Follower Examples
21:36 TiVo to Streaming
22:36 Blockbuster and Piracy
23:20 Netflix Pivot Lessons
24:13 Big Company Pivots
25:00 Apple Burns Ships
26:17 Microsoft Cloud Shift
27:20 Life Sciences vs Security
29:56 Agile Is Outdated
32:02 AI Coding Evolution
33:28 ADRs and Review Load
37:18 Code Rot Reality
40:01 Adopt AI or Fall Behind
43:11 Token Budgets and Local Models
44:12 153K Line PR Story
46:01 AI on the Periphery
47:17 Wrap and Cliffhanger

Didi & Lital are joined by their kids Ron and Leia, starting with reactions to the Jaylen Brown trade, which they argue reflects ownership-driven, analytics-obsessed decision-making that ignores entertainment value. The conversation shifts to how people measure the wrong outcomes, using a podcast on the stereotype “Asians are bad drivers,” where fatalities per mile show Asians have the lowest death rates, prompting debate over whether safety outcomes match everyday “bad driving” annoyances and traffic impacts. They then discuss Waymo and self-driving tech, noting Boston and New York resistance tied to unions and politics, while highlighting benefits for disabled and blind riders and claims that autonomous vehicles are significantly safer than humans. They compare tech disruption to containerization and self-checkout, discuss compromises like retraining and early retirement, and touch on public fear of rare autonomous crashes, liability, and lobbying.
 
Topics
01:01 Jaylen Brown Trade Shock
03:17 Analytics Versus Entertainment
06:38 Swift Kelce Wedding Rumor
09:26 Measuring the Wrong Things
10:02 Bad Drivers and Outcomes
15:53 Traffic Costs and Flow
19:00 Self Driving Cars Arrive
20:09 Boston Pushback on Waymo
22:08 Unions vs Waymo
23:33 Accessibility Case
24:43 Safety Data Debate
25:31 Teamsters Compromise
29:40 Automation Jobs Shift
33:24 Right to Drive
38:31 Control and Politics
40:33 Rare Crashes Liability
44:05 Lobbying and Wrap Up
45:38 Podcast Farewell

DiDi and Lital open with a summer sports update: the Red Sox are struggling, they discuss NBA and hockey player preferences for low- or no-income-tax states and argue Massachusetts’ millionaire tax affects teams’ ability to retain talent. Lital shares the atmosphere in downtown Boston during the Scotland soccer matches, with Scottish fans filling pubs and causing venues to run out of alcohol, plus light talk on global fandom and star players like Ronaldo, Messi, and Mbappé. The conversation shifts to business and product development: a request for dark theme triggers a broader point that it’s “not about the nail” or dark mode, but about organizational rate limits beyond coding, including uncertainty, fear, testing infrastructure, and release processes. They describe using AI to generate themes and emphasize automated testing, sandboxes, MVP scoping, and making faster business decisions while managing risk across a large customer base.
 
Topics
00:27 Red Sox Reality Check
01:08 Taxes and Team Trades
03:15 Hockey and No Tax States
05:20 Scottish Fans Take Boston
07:03 Modeling and Pub Planning
08:48 Globalization and Soccer Stars
12:18 From Sports to Business
12:42 Its Not About the Nail
14:13 Dark Theme Debate
18:28 AI Changes Product Delivery
20:53 Testing Infrastructure and MVP
29:46 Automation Over Manual QA
32:29 Shipping Faster in the Cloud
34:51 Wrap Up and Goodbye

Didi and Lital are joined by their kids Ron and Leia. The group jokes about family life and discusses Ron’s plan to attend RIT in Rochester to study computer engineering, partly as a smoother path for staying in the U.S. They run a round-robin of America-themed songs. The main discussion centers on self-agency versus powerlessness, contrasting internal vs. external locus of control, criticizing rule-following that discourages initiative, using examples like fire alarms, 9/11 messaging, and school shooter drills, and applying agency to work challenges such as AI, layoffs, and collaboration barriers by urging people to act, adapt, learn, or leave rather than complain.
The Day The Music Died: American Pie (documentary): https://www.paramountplus.com/movies/video/vrmcb4z5mbx3CwjVzC4L5x_76JIz1t8h/
 
Topics
00:43 Family Check In and Summer Plans
02:07 Ron Heads to RIT
04:27 Fourth Playlist Rules
05:27 Round One Song Picks
10:01 Round Two Curveballs
12:23 American Pie Deep Dive
14:25 Born in the USA Meaning
15:39 America the Beautiful Finale
16:49 New Topic Self Agency Debate
18:04 Defining Agency
19:55 Mud Ball Perspective
21:42 Culture Discourages Control
25:31 Rule Followers and Empire
26:59 Alarms and Shooter Drills
30:25 Haiti and Fatalism
33:40 Can Our Generation Succeed
35:32 Learning Disabilities Debate
39:20 Building Agency in Kids
41:09 Workplace Powerlessness
43:02 Ride the AI Wave
45:32 Freedom Needs Defending
46:43 Fourth of July Signoff

Didi and Lital discuss the brief restricted release and takedown of Anthropic’s Mythos/Fable, claims about a Glasswing participant issue, AWS researchers finding easy jailbreaks, and resulting U.S. government demands. They argue advanced AI can be treated like regulated technologies (encryption, bio materials) and propose licensing/KYC-style access for high-end models to enable research benefits (e.g., cancer immunotherapy advances) while limiting misuse.
 
Topics
00:20 Whistling Debate
01:32 Misophonia Confessions
02:34 Broadway Edition Begins
03:04 Cats and Memory
03:57 Phantom Deep Cut
05:37 West Side Story Picks
07:14 Fiddler and Broadway Roots
08:53 Porgy and Bess Choices
09:37 Grease Lightning Trivia
11:06 Over the Rainbow Origins
12:24 Rent and Chicago Finale
14:01 Broadway Song Tangent
14:28 Impossible Dream Memories
15:54 Mythos Release Fallout
17:11 How the Takedown Happened
22:21 AI as a Weapon History
30:33 Regulating Like Encryption
33:04 AI Boosting Cancer Therapy
39:08 Licensing Model Wrap Up
39:50 Closing and Next Episodes

DIdi, Lital and Leia open with a wide-ranging sports update (World Cup, the Patriots signing A.J. Brown and hopes for adding Stefon Diggs, Stanley Cup excitement, NBA comebacks, and critiques of Boston teams’ performance and decision-making). They then discuss Anthropic’s Mythos frontier model, which was delayed after consultation with the US government due to concerns it could rapidly uncover code vulnerabilities at scale; it was first released to select “Glasswing” organizations and later to the public with claimed restrictions. DD describes Mythos as highly autonomous, memory-enabled, and “token hungry,” able to perform extensive research and brute-force testing with minimal prompting. They warn of a “patching apocalypse” as vulnerability discovery rates surge, urging CISOs—especially in hospitals and infrastructure-heavy environments—to plan for continuous, cheap patching via redundancy, improved deployment architecture, and vendor coordination, and to consider cloud services like AWS to offload patching where possible.
 
Topics
00:34 Quick Sports Update
04:36 Boston Teams Reality Check
06:45 What Is Mythos
09:22 Mythos in Real Workflows
11:36 Memory Skills and Token Costs
13:54 Cyber Risk and Patch Apocalypse
15:28 How CISOs Should Prepare
20:45 Redundancy and Upgrade Architecture
22:33 Inevitable CVE Flood
25:26 Services and Jobs Shift
27:30 Plan Cloud and Re Architect
28:58 Wrap Up and Subscribe

DiDi and Lital welcome their daughter Leia back to the show to recap prom, graduation, and her upcoming move to Reichman University in Israel to study business administration and entrepreneurship, including details about her dorm choice and family help with tuition and a car. The hosts then run through a breakup-songs theme, naming picks such as U2’s “With or Without You,” Sinéad O’Connor’s “Nothing Compares 2 U,” the Greg Kihn Band’s “The Break-Up Song,” Scorpions’ “Still Loving You,” Alanis Morissette’s “You Oughta Know,” Carrie Underwood’s “Before He Cheats,” Hall & Oates, Phil Collins’ “Against All Odds,” Kelly Clarkson’s “Since U Been Gone,” multiple Taylor Swift songs, The Eagles, and Air Supply, and suggest a future part two. They shift to reflections on graduation, praising Leia’s “Excellence Award” and academic progress while contrasting it with her twin brother’s straight As and high SAT score but fewer recognitions and college acceptances, arguing schools and society reward hard work, relationships, and “participation” over outcomes and excellence, connecting this to hiring, sports team structure, AI’s impact on work, and the film “Broadcast News.”
 
Topics
00:50 Leia Returns
01:47 Prom and Graduation Recap
03:09 College Plans Israel
04:09 Dorms and Family Deal
06:47 Breakup Songs Theme
08:19 Classic Breakup Picks
10:20 Gen Z Taylor Era
12:09 80s and 90s Rage
16:23 Power Anthems Finale
18:06 Avril Lavigne Vibes
18:47 Taylor Swift Picks
19:36 Kids and Politics Talk
20:13 Eagles and Singalong
21:40 More Breakup Music Tease
22:44 Leia Graduation Wins
26:22 Ron and Merit Debate
30:14 Outcome Over Effort
32:31 Israel vs US Admissions
34:44 Hockey and A Players
41:15 AI Raises the Bar
43:02 Excellence and Curves
44:49 Movie Recommendation
45:52 Podcast Wrap Up

Didi and Lital open with casual sports talk (NBA, NHL, World Cup traffic) before returning to a cliffhanger about soaring AI token costs. They cite Marc Benioff saying Salesforce will pay Anthropic $300M in tokens for coding, and discuss Microsoft pausing Anthropic use after exhausting its AI budget. They argue costs are driven by energy/physics and by developers running many parallel agents, making outcome measurement crucial so token spend aligns with valuable work rather than performative usage. They describe organizational resistance ("Luddites") and poor prompting/problem decomposition ("one-liners"), stressing the need for core computer science thinking. They predict new solutions: on-prem/local models, company-built data centers, and software “broker” layers to route tasks to the right models and enforce policy. They describe building such a layer for security SOC triage to reduce alerts to ~2% needing humans, and discuss tradeoffs between general models for innovation and cheaper specialized models for repetitive tasks. They close noting rapid startup exits in this space.
 
Topics
00:24 Sports Rants
02:43 World Cup Traffic
03:43 AI Token Shock
07:59 Measuring AI Productivity
08:45 Luddites And One Liners
13:17 Family Travel Chaos
15:02 Token Optimization Layers
19:14 Generalists Vs Specialists
23:45 Efficiency War Stories
26:57 Future AI Cost Tools
28:55 Startup Exits Teaser
29:43 Wrap Up And Outro

DiDi and Lital discuss Elon Musk’s prediction that apps will be replaced by AI agents in 3–5 years, reviewing phone UI history (T9, BlackBerry, iPhone/App Store) and questioning future pricing models (tokens, bandwidth, on-device GPUs). They debate which apps persist (visual content, dating), how aggregators like OpenTable/Airbnb/booking.com may remain, risks of paid recommendations and transparency, and predict Google could dominate consumer agents while Apple lags. Producer Dave adds an Apple user perspective.
 
Topics
01:10 Summer Theme Kickoff
01:46 Summertime Classics Round One
05:03 Boys of Summer and ABBA Remix
07:37 Summer of 69 and Will Smith
09:19 Grease Nights and Van Halen
11:41 AI Returns No Apps Just Agents
12:36 Phone UI Evolution T9 to BlackBerry
15:15 BlackBerry To iPhone Shift
15:56 App Store And Bandwidth Money
16:48 Agents Change Interfaces
18:12 Audio Versus Visual UX
20:39 Portals Replaced By Agents
22:18 Paid Recommendations And Truth
23:31 Tokens Edge Compute Pricing
26:20 Aggregators Still Matter
31:53 Agent Optimization Opportunity
33:40 Google Versus Apple Prediction
36:11 Producer Dave Apple Perspective
37:25 Wrap Up And Podcast Outro