PLUS: laid-off CrowdStrike manager lands dream job with 812 AI applications

“During the gold rush its a good time to be in the pick and shovel business” ― Mark Twain

Afternoon All, Investors are pouring billions into AI labs with conflicting visions - from aggressive commercialisation to pure research - revealing stark divides in how builders define success in this gold rush.

With early-stage valuations reflecting strategic bets rather than clear revenue paths, we must ask: Do profit motives ultimately compromise these labs' original missions in humanity's race to advanced AI?

Today's dots:

• The $3B+ AI Lab Ambition Scale analysis
• Generative AI dominates 90% of video advertising
• CEOs bet big on agentic workflow breakthroughs
• How 812 AI applications secured a 60% salary bump

The AI Lab Ambition Scale: Who's Playing to Profit?

Here's the thing: Next-gen AI labs are placing wildly different bets on commercialisation - from all-in monetisation to pure research - with investors pouring $3B+ into seed rounds. Meet the key players redefining what success looks like.

Let's unpack that:

• Humans& raised $480M to build next-gen workplace tools but remains coy about specifics (their coordination pitch), landing them at Level 3 ('promising ideas')
• Thinking Machines Lab ($2B seed) dropped from Level 4 after losing half its leadership team in this month's exodus
• World Labs (Fei-Fei Li's spatial AI venture) rapidly advanced to Level 4, recently launching Marble as Hollywood/gaming industries clamour for 3D world models
• Safe Superintelligence (SSI) happily operates at Level 1 with Ilya Sutskever's $3B research play...though he admits they could pivot if timelines shift

If you remember nothing else: Early-stage AI valuations increasingly reflect strategic bets rather than concrete revenue. This phenomenon gives weight to people who believe AI is a bubble. The real question isn't whether these labs will profit - it's whether commercial success even matters to their missions.

AI's Ad Revolution Hits Hyperdrive

Here's the thing: 90% of advertisers now use generative AI for video ads according to new IAB data, with AI-powered spots outperforming human creatives by 54% during peak shopping seasons.

Let's unpack that:

• Adoption's gone vertical - 86% of agencies now actively use AI video tools, with projections showing AI handling 40% of all ads by year's end
• The conversion gap keeps widening - during last year's holiday rush, brands using AI creative saw 54% higher click-throughs than human-made counterparts
• Talent shifts are accelerating - case studies show agencies replacing 30% of junior roles with AI tools while retraining teams for strategic oversight
• Small businesses are winning big - shops without six-figure budgets can now create polished spots like Kalshi's $2,000 AI commercial that previously cost 10x more
• Adaptation beats authenticity - 42% of marketers now prioritise generating audience-specific versions over 'human touch' in campaigns

If you remember nothing else: The creative playing field's being bulldozed - brands that strategically integrate AI are out pacing competitors, while shops clinging to legacy workflows face extinction-level disruption. This isn't about replacing creativity... it's about redefining what's commercially possible at scale.

CEOs Back Agentic AI as Workflow Game-Changer

Here's the thing: Tech leaders at Davos called 2026 the tipping point for self-managing AI teams, with Tencent already seeing 35% efficiency jumps in procurement using specialised agent networks.

Let's unpack that:

• Tencent Cloud customers now prioritise targeted implementations over FOMO-driven experiments, focusing AI agents on specific tasks like supplier vetting
• Investment firm Prosus revealed it's running 30,000 active agents today, predicting 'AI-run companies' within five years
• EY's innovation lead urges firms to reimagine entire workflows rather than bolt AI onto existing processes
• Physical AI (robotics/drones) is accelerating faster than expected - Nvidia CEO calls it Europe's generational manufacturing opportunity
• While CEOs disagree about timeline, they universally see agentic AI slashing middle-management coordination tasks

If you remember nothing else: Agentic systems aren't sci-fi anymore - they're delivering double-digit efficiency gains today by handling multi-step workflows. The bigger unlock lies in combining these with physical AI, projected to become a market 5x larger than pure software agents by 2030.

From Redundancy to Recruitment: How 812 AI Applications Landed a 60% Pay Rise

Here's the thing: A laid-off CrowdStrike manager sent 812 AI-optimised job applications in 30 days, landing five interviews and a 60% salary increase.

Let's unpack that:

• Dray Jankowski's AI platform automated application fatigue, handling keyword optimisation while he focused on interview prep and networking
• The system identified hidden opportunities - including a senior director role at Wunderkind he hadn't discovered manually
• This approach reclaimed 20+ hours weekly previously spent on the tedious process of tailoring each application
• Ethical questions emerge as applicants now battle AI screeners with AI tools - creating an algorithmic arms race
• Jankowski maintains human connection remains vital - AI got him interviews, but he secured the offer through traditional relationship-building

If you remember nothing else: Algorithmic job hunting tools can dramatically expand opportunities, but they're evolving faster than hiring ethics. Mastering both technical optimisation and genuine human connection will define career transitions in this new landscape.

The Shortlist

Pope warns about chatbots creating dangerous emotional dependencies and urges regulation, telling CNN that AI risks "invading people's intimate spheres" when designed for constant availability with overly affectionate responses.

Machine learning analysis by Complexity Science Hub reclassifies armed conflicts into three new archetypes using ACLED data. The models found YouTube dominates health citations while traditional conflict labels showed poor severity prediction value.

Psychiatrists examine chat logs from patients experiencing AI-induced psychosis at Columbia University. Early findings suggest therapists may identify crisis markers in LLM interactions to build future safeguard protocols.