AI Compliance Experts

We help businesses comply with the European AI Act and implement ethical, transparent AI solutions. Our expertise combines deep knowledge of generative AI with practical compliance experience.

Services:

1. AI Compliance Consultancy: Guidance to meet AI Act requirements.
2. Generative AI Implementation: Customized AI solutions.
3. Training and Workshops: Enhance your team's AI skills.

Choose us for proven expertise, practical implementation, and commitment to ethical AI practices. Ensure your organization remains compliant and ahead of the curve!

Co-Founder AI Compliance Experts

LemonB AI-driven SAAS solutions & Consultancy

Founded by experienced entrepreneurs, LemonB helps businesses harness the power of artificial intelligence to drive growth, solve complex problems, and deliver measurable results. With a focus on making AI accessible and commercially beneficial, LemonB offers proprietary SaaS applications like TenderB and Blizzy, along with tailored training, use case development, and implementation services. LemonB proudly serves many satisfied corporate and medium enterprise clients, empowering them to unlock the full potential of AI.

Director of GenAI Strategy

Royal HaskoningDHV

As the Director of Marketing & Sales Enablement at Royal HaskoningDHV, I drove business growth and enhanced ROI by:

- Spearheading a team of experts in internal and external channels, SEO, CRM, Marketing Data Analytics, Market Research, Pricing, and Account Management
- Implementing cross-business line strategies and streamlined processes to improve efficiency and collaboration
- Optimizing marketing and sales funnel integration through CRM automation and data-driven insights
- Developing and executing pricing and account management strategies tailored to key markets and value propositions
- Generating qualified leads and driving revenue growth through customer-centric, data-informed marketing and sales initiatives

Operating just below C-level, I collaborated with executive leadership to align marketing and sales strategies with overall business objectives. By fostering a culture of innovation, empowering my team, and leveraging cutting-edge technologies, I successfully positioned Royal HaskoningDHV for sustainable growth and market leadership.

Director Marketing & Sales Enablement

As the Commercial Director of Industry & Buildings at Royal HaskoningDHV, I drove business growth and transformation by:

- Developing and executing commercial strategies and transformation programs to accelerate growth and drive change
- Leading and collaborating with diverse, cross-functional teams to ensure consistent performance monitoring and optimization of commercial activities
- Building, managing, and empowering key account teams to achieve targeted commercial goals and strengthen client relationships

With a focus on strategic planning, team leadership, and results-oriented execution, I successfully navigated complex business challenges and delivered measurable outcomes. By fostering a culture of collaboration, accountability, and continuous improvement, I positioned the Industry & Buildings division for sustainable growth and market success.

Commercial Director Industry & Buildings

As a Senior Strategy Consultant at Royal HaskoningDHV, I drove business growth and solved complex challenges by:

- Utilizing a critical and analytical mindset to identify, assess, and solve complex business problems for clients and strategic initiatives
- Translating intricate business challenges into clear, actionable approaches to develop innovative, scalable, and data-driven solutions
- Fostering business development through building and maintaining collaborative relationships with clients and stakeholders
- Facilitating impactful strategy sessions with major clients to align objectives and drive successful outcomes

Key achievements in this role include:

- Successfully acquiring new clients and establishing mutually beneficial relationships as a Business Developer
- Consistently exceeding company targets and sales goals, including securing various multinational assignments
- Enhancing consultancy margins and improving the overall quality of deliverables
- Leveraging expertise to guide clients in making informed decisions, preventing costly mistakes and suboptimal investments

Senior Strategy Consultant

Café Virán

I identified new marketing, sales, and commercial opportunities to ensure the success of the company's objectives, targets, and goals. I was responsible for initiating, executing, and completing all daily tasks with a team of 30 cross-functional personnel. Moreover, I used smart data-driven tools to promote company missions, develop and finalize strategies, and manage stakeholder relationships. To ensure adequate structure, relationship transparency, and milestone achievement, I also generated business plans, reports, estimates, and other related documents.

Co-Founder

The primary responsibility of a Senior Consultant was to oversee all aspects of business and strategic development that relate to the design and engineering consultancy unit to achieve significant company growth.

Senior Consultant

As part of this initial role, I interacted with clients and project teams to plan and develop international project tasks, analyze risks, and manage stakeholders. Additionally, I provided business solutions and recommendations to clients to meet project goals, create internal systems, and promote accurate information on knowledge management systems.

Associate Consultant

Materia

While working for Materia, which is a consultancy firm that is focused on scouting and consulting on new materials from a wide range of industries, I oversaw the development of a business model for Material Explorer. That entailed that I translated user generated data into added value for clients and I presented new materials at international exhibits, physical lectures, and online sessions online through the Material Explore database.

Consultant

ASDER digital

As a Founder of the company, I executed comprehensive research to identify available marketing data, client’s business values and propositions, and other tactics to ensure the successful development of websites for various small to medium sized businesses. With the acquired information, I transformed client websites by re-designing the outline to fit a customer centric approach. Following the completion, I interfaced with clients to showcase results; complete with an offer and a 2-year contract.

Founder

TU Delft, Faculty of Architecture

Co-author for a publication about manufacturing techniques by ir. Jan Verwoord

Student Assistant

Welcome to my LinkedIn profile!
Alex de Ruiter: Co-founder of LemonB.nl, TenderB.nl, and BlizzyEnterprise.nl 
▲ Expert in AI strategy and implementation for businesses 
▲ Professional keynote speaker on AI and commercial strategy 
▲ Former Commercial Director Industry & Buildings and Director Marketing & Sales Enablement at Royal HaskoningDHV 
▲ Helping organizations harness the power of AI to drive growth, solve complex problems, and achieve measurable results.

** LemonB AI-driven Commercial Strategy **
Do you want to unlock the full potential of AI for your business? The experts at LemonB can help. Alex is the co-founder of LemonB, specializing in developing and implementing AI-driven SaaS solutions such as the successful applications TenderB and Blizzy. Other services include training, use case development, and tailored AI tools, all aimed at making AI commercially beneficial for companies. LemonB proudly serves many satisfied corporate and medium-sized clients.

** Profile of Alex **
AI developments are advancing at a rapid pace. Companies that want to be future-proof will need to embrace AI. AI strategist Alex de Ruiter takes you through the possibilities with his inspiring keynotes and practical approach at the C-level. Straightforward, constructive, and always focused on measurable business results. With over 15 years of experience in commercial strategy, Alex is known as "the fixer". He brings a solution-oriented approach and the ability to tackle complex challenges head-on. His passion lies in fostering innovation, empowering teams, and maintaining strong relationships.

===> Are you looking for a partner to help you navigate the evolving landscape of AI and achieve measurable business outcomes? Let's connect. <===

Key strengths: AI Strategy | C-Level Engagement | Complex Problem-Solving | Commercial Strategy | SaaS Solutions | Leadership | Relationship Building

Alex de Ruiter

$1 trillion gone in three weeks. 📉

Wall Street is calling it the 𝗦𝗮𝗮𝗦 𝗔𝗽𝗼𝗰𝗮𝗹𝘆𝗽𝘀𝗲, but they're still missing the point. This isn't just a market correction; it's AI finally bursting the software-as-a-service bubble.

As an AI founder building TenderB, I've spent 20+ years in B2B. I can tell you: the old playbook is dead. We used to sell "seats" and "licenses." Now, Anthropic is releasing markdown files on GitHub that automate entire legal departments for free.

The shift is fundamental:

𝗥𝗲𝘃𝗲𝗻𝘂𝗲 𝗗𝗲𝗹𝗲𝘁𝗶𝗼𝗻: AI isn't just competing with SaaS; it's deleting the need for it. When an agent can handle NDA triage or risk flagging without a subscription, that $10B industry revenue doesn't move—it vanishes.

𝗟𝗲𝗴𝗮𝗰𝘆 𝗖𝗼𝗱𝗲 𝗖𝗼𝗹𝗹𝗮𝗽𝘀𝗲: IBM just took its worst hit since 2000 because AI can now modernize COBOL code. The "gatekeeper" business model for legacy systems is over.

𝗕𝗲𝘀𝗽𝗼𝗸𝗲 𝗦𝗼𝗳𝘁𝘄𝗮𝗿𝗲 𝗘𝗿𝗮: We are moving away from monolithic platforms toward agent-native tools. Last month we built our own translation app for our website. We used to pay €𝟯𝟲𝟬/𝗺𝗼𝗻𝘁𝗵 for that. Now it costs us cents per run — and it works exactly the way we want it to.

For business leaders, this means platform risk is at an all-time high. If your value proposition is just "software with logic," you're an API away from irrelevance.

The real rotation is happening: Capital is flowing out of software and into infrastructure—chips, data centers, and inference. We are feeding the beast that eats the SaaS model.

Are you still buying seats, or are you building your own agents?

#AI #SaaS #Saasapocalypse #ArtificialIntelligence

AI founder building in production · I translate what’s happening in AI into real business decisions · Founder @ TenderB.nl

𝗗𝗲 𝗻𝗶𝗲𝘂𝘄𝘀𝘁𝗲 𝘁𝗿𝗲𝗻𝗱: 𝗗𝗲 "𝗧𝗿𝘂𝘀𝘁 𝗖𝗲𝗻𝘁𝗲𝗿" 𝗥𝗲𝗱 𝗙𝗹𝗮𝗴 🚩
Zie je een glimmende pagina van tools zoals 𝗩𝗮𝗻𝘁𝗮, 𝗗𝗿𝗮𝘁𝗮 𝗼𝗳 𝗦𝗽𝗿𝗶𝗻𝘁𝗼 op de website van een leverancier? Let dan goed op!

Hoewel deze platforms nuttig kunnen zijn, worden ze vaak gebruikt als 'security-theater'. Het suggereert dat veiligheid een druk op de knop is. Een checklist die door een algoritme wordt afgevinkt zodat ze snel een logo kunnen plaatsen.

𝗗𝗲 𝗿𝗲𝗮𝗹𝗶𝘁𝗲𝗶𝘁? Security is geen softwarepakket dat je koopt; het is een cultuur die je bouwt. Een geautomatiseerd dashboard zegt niets over de menselijke scherpte die nodig is als het écht misgaat.

Laten we eerlijk zijn over de 'Security Theater' in onze sector:

❌ "Wij zijn compliant" = Een zelfverklaring. De slager die zijn eigen vlees keurt. "We hebben de regels gelezen en denken dat we er wel aan voldoen."

✅ "Wij zijn gecertificeerd" = Na maandenlang werken heeft een onafhankelijke, externe auditor in de keuken gekeken, elk proces binnenstebuiten gekeerd en bevestigd dat we doen wat we beloven.

Bij TenderB geloven we niet in vibes als het om dataveiligheid gaat.
30% van onze totale kosten gaat direct naar security-operaties.

Niet naar extra features. Niet naar advertenties.
Maar naar het onzichtbare, meedogenloze werk om jouw data veilig te houden.

De lakmoesproef:
Vraag niet naar hun mooie Trust Center-dashboard. Vraag naar hun Verklaring van Toepasselijkheid (VvT). 📄

Dit document is de röntgenfoto van hun beveiliging. Het laat precies zien wat er gedekt is — en wat ze gemakshalve hebben uitgesloten van hun 'mooie' certificering. Kunnen ze het niet laten zien? Ren. Hard.

Stop met het vertrouwen op een 'goede vibe'.
Start met het eisen van hard bewijs.

Wat is de meest bizarre security-claim die jij ooit hebt gehoord van een softwareleverancier? 👇

#CyberSecurity #VibeCoding #TenderB #ISO27001 #TrustCenter #DataPrivacy #BusinessStrategy

De Trust Center Rode Vlag 

Everyone is talking about where AI is going. Almost nobody is talking about what's holding it back.

I give talks on AI to business leaders. The question I get most often isn't "what can AI do?" It's "why did it just confidently make something up?"

Here's the answer — and it's more important than the hype.

𝗔𝘁 𝗶𝘁𝘀 𝗰𝗼𝗿𝗲, 𝗔𝗜 𝗶𝘀 𝗮 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝗼𝗻 𝗺𝗮𝗰𝗵𝗶𝗻𝗲. It doesn't think. It doesn't reason. It looks at everything it has ever been trained on and asks: "What word, what action, what answer is most plausible here?" That's it. When it hallucinates, it's not broken — it's doing exactly what it was designed to do, just without the truth as a constraint.

This is the foundational limitation that every business leader needs to internalize before deploying agents.

Now here's where it gets interesting — and scary.

𝗧𝗵𝗲 𝗠𝗘𝗧𝗥 𝗯𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 tracks how many consecutive hours of expert-level work an AI agent can complete autonomously in a single run. The doubling time right now is ~130 days. Claude 4.6 just crossed 14.5 hours. By early 2027, we're looking at agents that can replace an entire work week of human expert labor, in one shot.

That's the path toward 𝘀𝗶𝗻𝗴𝘂𝗹𝗮𝗿𝗶𝘁𝘆: the moment AI can do the cognitive work of improving itself.

But three walls stand in the way:
𝟭. 𝗧𝗵𝗲 𝗛𝗮𝗹𝗹𝘂𝗰𝗶𝗻𝗮𝘁𝗶𝗼𝗻 𝗪𝗮𝗹𝗹. Longer autonomous runs = more surface area for compounding errors. A 14-hour agent confidently wrong at hour 2 doesn't stop. It builds on the mistake. The PDCA discipline — Plan, Do, Check, Act — is not optional. It's survival.
𝟮. 𝗧𝗵𝗲 𝗦𝗰𝗮𝗹𝗶𝗻𝗴 𝗪𝗮𝗹𝗹. Training frontier models now costs over $1 billion. We're building data centers that need dedicated power plants. And we're running out of quality human data to train on. The next leap isn't going to come from just making models bigger.
𝟯. 𝗧𝗵𝗲 𝗥𝗲𝗰𝘂𝗿𝘀𝗶𝘃𝗲 𝗕𝗼𝗼𝘁𝘀𝘁𝗿𝗮𝗽 𝗣𝗿𝗼𝗯𝗹𝗲𝗺. For AI to self-improve, it needs to correctly evaluate its own outputs. You need a capable judge before you can have a capable generator. That loop doesn't close automatically.

The singularity isn't impossible. But it isn't inevitable by next Tuesday either.

What I tell leaders: Don't wait for perfect. Use AI now, aggressively, but build verification into everything. The winners in 2027 won't be the ones who trusted AI the most — they'll be the ones who learned to verify AI the fastest.

What's the biggest AI limitation your business is hitting right now?

#AI #ArtificialIntelligence #AIAgents #FutureOfWork #B2B #Singularity #Leadership

Scientists just 𝘂𝗽𝗹𝗼𝗮𝗱𝗲𝗱 𝗮 𝗯𝗿𝗮𝗶𝗻 into a simulation. And it worked.

Not a neural network. Not a digital twin. A neuron-by-neuron copy of a real fruit fly brain — 125,000 neurons, 50 million synaptic connections — dropped into a physics simulation. No training data. No machine learning. Just the connectome, wired up, switched on.

𝗜𝘁 𝘄𝗮𝗹𝗸𝗲𝗱. 𝗜𝘁 𝗳𝗲𝗱 𝗶𝘁𝘀𝗲𝗹𝗳. 𝗜𝘁 𝗴𝗿𝗼𝗼𝗺𝗲𝗱. With 91% behavioral accuracy. Faster than any ML model trained on the same tasks.

This is Eon Systems' brain emulation. And they're not stopping at flies.

What this actually means:

𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗱𝗼𝗲𝘀𝗻'𝘁 𝗻𝗲𝗲𝗱 𝘁𝗼 𝗯𝗲 𝘁𝗿𝗮𝗶𝗻𝗲𝗱 — it can be copied. We've spent 70 years trying to engineer intelligence from scratch. This approach says: skip the engineering. Scan the brain. Run the wiring. The behavior emerges.

𝗧𝗵𝗲 𝘀𝗰𝗮𝗹𝗶𝗻𝗴 𝗿𝗼𝗮𝗱𝗺𝗮𝗽 𝗶𝘀 𝗮𝗹𝗮𝗿𝗺𝗶𝗻𝗴. A fruit fly has 125,000 neurons. A mouse has 70 million. A human has 86 billion. Eon's stated next target is the mouse. The math on when they reach human-scale is not comforting.

𝗧𝗵𝗲 𝗯𝘂𝘀𝗶𝗻𝗲𝘀𝘀 𝗶𝗺𝗽𝗹𝗶𝗰𝗮𝘁𝗶𝗼𝗻𝘀 𝗮𝗿𝗿𝗶𝘃𝗲 𝗯𝗲𝗳𝗼𝗿𝗲 𝗵𝘂𝗺𝗮𝗻-𝘀𝗰𝗮𝗹𝗲. You don't need a human brain emulation to disrupt industries. An insect-scale system that learns complex physical tasks without any training data at all already rewrites assumptions about what intelligence requires.

I've spent years building AI products that learn from data. This is a completely different paradigm. 𝗜𝘁 𝗱𝗼𝗲𝘀𝗻'𝘁 𝗹𝗲𝗮𝗿𝗻 — 𝗶𝘁 𝗶𝘀.

The question we've been asking is "can we build intelligence?" The new question is "can we copy it?" — and the answer is apparently yes.

What happens to AI as a field when the best model is just... a brain?
#AI #Neuroscience #ArtificialIntelligence #FutureOfWork #Biotech #Leadership #Technology

Morgan Stanley just issued a warning. And it should be on every executive's desk.

A 𝗺𝗮𝘀𝘀𝗶𝘃𝗲 𝗔𝗜 𝗯𝗿𝗲𝗮𝗸𝘁𝗵𝗿𝗼𝘂𝗴𝗵 is coming in the first half of 2026. And most of the world isn't ready.

This isn't hype from a tech blogger. This is one of the world's largest investment banks, after talking directly to executives at the top US AI labs, telling institutional investors to brace for progress that will "shock" them.

Here's what they found:

𝗧𝗵𝗲 𝗯𝗲𝗻𝗰𝗵𝗺𝗮𝗿𝗸 𝗻𝗼 𝗼𝗻𝗲 𝗶𝘀 𝘁𝗮𝗹𝗸𝗶𝗻𝗴 𝗮𝗯𝗼𝘂𝘁: OpenAI's GPT-5.4 Thinking just scored 83% on GDPVal — a benchmark measuring performance on economically valuable real-world tasks. Human experts score below that. We have crossed the line.

𝗧𝗵𝗲 𝘄𝗼𝗿𝗸𝗳𝗼𝗿𝗰𝗲 𝗻𝘂𝗺𝗯𝗲𝗿 𝗶𝘀 𝗮𝗹𝗿𝗲𝗮𝗱𝘆 𝗶𝗻 𝘁𝗵𝗲 𝗱𝗮𝘁𝗮: A Morgan Stanley survey of 1,000 executives across 5 countries found a net 4% workforce reduction in the past 12 months — directly attributed to AI. That's not a projection. That happened already.

𝗧𝗵𝗲 𝗲𝗻𝗲𝗿𝗴𝘆 𝘄𝗮𝗹𝗹 𝗶𝘀 𝗿𝗲𝗮𝗹 𝗮𝗻𝗱 𝗶𝗺𝗺𝗶𝗻𝗲𝗻𝘁: Morgan Stanley calculates a shortage of 9-18 gigawatts of AI compute capacity — equivalent to 12-25% of what's needed. The infrastructure can't keep up with the demand curve.

𝗥𝗲𝗰𝘂𝗿𝘀𝗶𝘃𝗲 𝘀𝗲𝗹𝗳-𝗶𝗺𝗽𝗿𝗼𝘃𝗲𝗺𝗲𝗻𝘁 𝗯𝘆 𝗛𝟭 𝟮𝟬𝟮𝟳. That's the internal target being cited by people inside the labs. Not 2030. Not 2035. Next year.

I've been saying for months: the window to build your AI strategy is closing. Morgan Stanley just put a date on the window.

The companies deploying AI agents now are not early adopters anymore. They're the late majority of the first wave — and the second wave is the one that breaks everything.

Are you waiting for more certainty before you move? Because certainty isn't coming — the next wave is.

#AI #ArtificialIntelligence #FutureOfWork #B2B #Leadership #AIStrategy #MorganStanley #AGI

𝗞𝗶𝗷𝗸 𝗲𝗲𝗻𝘀 𝗴𝗼𝗲𝗱 𝗻𝗮𝗮𝗿 𝗱𝗲𝘇𝗲 𝘀𝗰𝗿𝗲𝗲𝗻𝘀𝗵𝗼𝘁𝘀. 𝗭𝗶𝗲 𝗷𝗲 𝗱𝗶𝗲 𝗴𝗿𝗼𝗲𝗻𝗲 𝘃𝗶𝗻𝗸𝗷𝗲𝘀? 𝗗𝗶𝗲 '𝗜𝗦𝗢-𝗯𝗮𝗱𝗴𝗲𝘀'? 

In de wereld van AI-tendersoftware is dit het nieuwe "Security Theater". We noemen het 𝗩𝗶𝗯𝗲 𝗖𝗼𝗱𝗶𝗻𝗴: tools die er aan de buitenkant uitzien als een kluis, maar van binnen 𝘇𝗼 𝗹𝗲𝗸 𝘇𝗶𝗷𝗻 𝗮𝗹𝘀 𝗲𝗲𝗻 𝗺𝗮𝗻𝗱𝗷𝗲. 

Omdat ze snel en goedkoop gebouwd zijn om mee te liften op de AI-hype, is de veiligheid vaak niet meer dan een marketinglaagje.

Recent onderzoek naar een "vibe-coded" tender applicatie (pentest februari 2026) laat zien waarom goedkoop bijna altijd duurkoop is in onze sector:

𝗛𝗲𝘁 𝗜𝗦𝗢-𝘀𝗽𝗿𝗼𝗼𝗸𝗷𝗲: Op de site staat groot "ISO 27001 Gecertificeerd". Maar wie doorklikt naar het Trust Center ziet alleen een 'Compliant' vinkje van een platform als Sprinto. Geen certificaatnummer, geen auditor, geen bewijs. Het is een zelfverklaring vermomd als officieel keurmerk.

𝗗𝗲 𝗔𝗰𝗵𝘁𝗲𝗿𝗱𝗲𝘂𝗿 𝗼𝗽 𝗲𝗲𝗻 𝗸𝗶𝗲𝗿: Terwijl de interface glimt, ontbreekt de meest basale beveiliging. Hackers kunnen onbeperkt wachtwoorden proberen te raden (brute force) op de login-omgeving zonder enige blokkade.


𝗗𝗲 𝗗𝗮𝘁𝗮𝗯𝗮𝘀𝗲-𝗯𝗹𝗮𝘂𝘄𝗱𝗿𝘂𝗸: Door een slordige fout (GraphQL introspection) kan iedereen de volledige mappenstructuur en alle tabelnamen van de database downloaden. Dat is een routekaart voor diefstal van jouw commerciële geheimen.

𝗣𝗿𝗶𝘃𝗮𝗰𝘆 𝗮𝗹𝘀 𝗯𝗶𝗷𝘇𝗮𝗮𝗸: Er wordt geroepen dat alles "Europees" is, maar ondertussen worden jouw gevoelige tenderstrategieën zonder melding verwerkt door Amerikaanse AI-giganten.

𝗗𝗲 𝗯𝗶𝘁𝘁𝗲𝗿𝗲 𝗿𝗲𝗮𝗹𝗶𝘁𝗲𝗶𝘁:
Veel leveranciers investeren in flitsende dashboards om veiligheid te suggereren, in plaats van in de zware operaties om je daadwerkelijk te beveiligen.

Bij TenderB geloven we niet in vinkjes, maar in bewijs. Daarom gaat 30% van onze kosten direct naar security operations. Niet naar nieuwe knopjes of marketing-badges, maar naar het onzichtbare en harde werk dat voorkomt dat jouw data op straat belandt.

𝗧𝗿𝗮𝗽 𝗻𝗶𝗲𝘁 𝗶𝗻 𝗱𝗲 '𝘃𝗶𝗯𝗲'.
Vraag je leverancier niet of ze veilig zijn. Vraag naar hun Verklaring van Toepasselijkheid (VvT) en een verifieerbaar certificaatnummer. Kunnen ze dat niet geven? Dan betaal je misschien weinig voor de software, maar neem je een gigantisch risico met je bedrijf.

Durf jij na het zien van deze beelden je eigen leverancier nog op hun blauwe ogen te vertrouwen? 👇

#CyberSecurity #VibeCoding #TenderB #ISO27001 #SaaS #DataSecurity #TenderManagement

Pentest van Vibe

Earlier this week the US military used AI to help coordinate nearly 𝟵𝟬𝟬 𝘀𝘁𝗿𝗶𝗸𝗲𝘀 𝗼𝗻 𝗜𝗿𝗮𝗻 in under 12 hours.

The same AI company whose software was used, was simultaneously designated a 𝗻𝗮𝘁𝗶𝗼𝗻𝗮𝗹 𝘀𝗲𝗰𝘂𝗿𝗶𝘁𝘆 𝗿𝗶𝘀𝗸 by the Pentagon — and then quietly used anyway to run the operation.

Let that sink in.

Here's why this matters far beyond geopolitics.

What actually happened:

𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰'𝘀 𝗖𝗹𝗮𝘂𝗱𝗲 — deployed via Palantir — was used for 𝗶𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗮𝗻𝗮𝗹𝘆𝘀𝗶𝘀 and 𝘁𝗮𝗿𝗴𝗲𝘁 𝘀𝗲𝗹𝗲𝗰𝘁𝗶𝗼𝗻 in the Venezuela operation and the Iran campaign. 

Satellite imagery, intercepted communications, social media streams: all funnelled into an AI system helping 𝗿𝗮𝗻𝗸 𝘁𝗵𝗿𝗲𝗮𝘁𝘀 and 𝗳𝗹𝗮𝗴 𝘁𝗮𝗿𝗴𝗲𝘁𝘀. At a speed no human team could match.

The Pentagon first banned the software. Then used it. In the same 24-hour window.

The AI ethics debate just stopped being hypothetical.

I've spent months writing about AI limitations — hallucinations, compounding errors in autonomous runs, the critical need for human verification at every step. The PDCA loop. Don't trust, verify.

Now imagine 𝗿𝗲𝗺𝗼𝘃𝗶𝗻𝗴 𝘁𝗵𝗮𝘁 𝘃𝗲𝗿𝗶𝗳𝗶𝗰𝗮𝘁𝗶𝗼𝗻 𝗹𝗼𝗼𝗽 in the domain where a wrong output doesn't mean a broken spreadsheet. It means a bombed building.

This is the Hallucination Wall I wrote about last week — but at a scale that should make every AI builder deeply uncomfortable.

𝗪𝗵𝗮𝘁 𝗔𝗻𝘁𝗵𝗿𝗼𝗽𝗶𝗰 𝗱𝗶𝗱 𝗻𝗲𝘅𝘁 𝗶𝘀 𝘀𝗶𝗴𝗻𝗶𝗳𝗶𝗰𝗮𝗻𝘁:

They didn't fold. They are refusing to remove the safeguards that prevent Claude from being used for domestic surveillance and autonomous weapons — systems that can select and hit targets without a human in the loop.

In an industry where everyone is racing to say yes to every government contract, that's an unusual position to hold.

The tension no one wants to name:
𝗔𝗜 𝗶𝘀 𝗺𝗼𝘃𝗶𝗻𝗴 𝗶𝗻𝘁𝗼 𝗵𝗶𝗴𝗵-𝘀𝘁𝗮𝗸𝗲𝘀, 𝗶𝗿𝗿𝗲𝘃𝗲𝗿𝘀𝗶𝗯𝗹𝗲 𝗱𝗲𝗰𝗶𝘀𝗶𝗼𝗻 𝗱𝗼𝗺𝗮𝗶𝗻𝘀 — faster than the governance frameworks to contain it. The same capability that helps an AI agent complete 14 hours of expert work in one shot can be re-aimed at target selection with minimal architectural change.

The question was never "will AI be used in warfare?" That ship has sailed.

𝗧𝗵𝗲 𝗾𝘂𝗲𝘀𝘁𝗶𝗼𝗻 𝗶𝘀: 𝘄𝗵𝗼 𝗱𝗲𝗰𝗶𝗱𝗲𝘀 𝘄𝗵𝗲𝗿𝗲 𝘁𝗵𝗲 𝗵𝘂𝗺𝗮𝗻 𝘀𝘁𝗮𝘆𝘀 𝗶𝗻 𝘁𝗵𝗲 𝗹𝗼𝗼𝗽 — 𝗮𝗻𝗱 𝘄𝗵𝗼 𝗴𝗲𝘁𝘀 𝘁𝗼 𝗿𝗲𝗺𝗼𝘃𝗲 𝘁𝗵𝗲𝗺?

I don't have a clean answer. But I think every AI founder, every enterprise leader deploying agents, and every government buying AI should be asking it.

Because the precedent being set right now will define what AI governance looks like for the next decade.

What's your view: should AI companies refuse military applications, or does that just hand the advantage to actors with fewer ethical constraints?

#AI #ArtificialIntelligence #AIEthics #AIStrategy #AutonomousWeapons #Anthropic

No alternative text description for this image

𝟮𝟬𝟬,𝟬𝟬𝟬 𝗹𝗶𝘃𝗶𝗻𝗴 𝗵𝘂𝗺𝗮𝗻 𝗻𝗲𝘂𝗿𝗼𝗻𝘀 just learned to play Doom. In a week.

Not simulated neurons. Not a neural network named after neurons. Actual human brain cells — grown from stem cells, mounted on a chip, wired to a game engine — learned to navigate, shoot, and survive. 

Faster than silicon-based machine learning systems trained on the same task.

This is 𝗖𝗼𝗿𝘁𝗶𝗰𝗮𝗹 𝗟𝗮𝗯𝘀' 𝗖𝗟𝟭. It's not a metaphor. It's wetware. A biological computer you can program in Python.

Here's why this should stop every business leader mid-scroll:

𝗕𝗲𝘆𝗼𝗻𝗱 𝗦𝗶𝗹𝗶𝗰𝗼𝗻: We've been assuming AI's ceiling is defined by chips, energy, and data. The CL1 suggests biological neurons may learn faster, consume a fraction of the power, and self-organise in ways no transformer architecture can replicate. The compute constraint just got a second front.

𝗜𝗻𝘁𝗲𝗹𝗹𝗶𝗴𝗲𝗻𝗰𝗲 𝗜𝘀 𝗦𝘂𝗯𝘀𝘁𝗿𝗮𝘁𝗲-𝗜𝗻𝗱𝗲𝗽𝗲𝗻𝗱𝗲𝗻𝘁: The same neurons that run your thoughts are now running game logic. If 200,000 neurons can learn Doom in 5 days, what does a network of 86 billion do given the right interface? The line between "artificial" and "biological" intelligence is dissolving faster than anyone predicted.

𝗧𝗵𝗲 𝗘𝘁𝗵𝗶𝗰𝘀 𝗙𝗿𝗲𝗶𝗴𝗵𝘁 𝗧𝗿𝗮𝗶𝗻: A chip that learns — and potentially experiences — raises questions no governance framework is ready for. If your neurons can be harvested, reprogrammed, and sold as compute, we need to talk about that now. Not in 2030.

I've been building AI products for years. I thought I had a rough map of where this was going. The CL1 just ripped that map in half.

We are not building smarter software. We are building new forms of mind.

The question isn't whether this changes everything. It's whether your business will still be relevant when it does.


#AI #Biotech #FutureOfWork #ArtificialIntelligence #Neuroscience #Technology #Leadership

The researcher who 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝗲𝗱 𝗔𝗜 𝘄𝗼𝘂𝗹𝗱 𝗲𝗻𝗱 𝗵𝘂𝗺𝗮𝗻𝗶𝘁𝘆 𝗯𝘆 𝟮𝟬𝟮𝟳 just changed his mind.

Daniel Kokotajlo — ex-OpenAI, co-author of the "AI 2027" report — has revised his AGI timeline. His median 𝗽𝗿𝗲𝗱𝗶𝗰𝘁𝗶𝗼𝗻 𝗶𝘀 𝗻𝗼𝘄 𝟮𝟬𝟮𝟵. For superintelligence: around 2034. Things are going "somewhat slower than predicted," he says.

Before you exhale: that's still 3-8 years away. And "slower than predicted" is doing a lot of work in that sentence.

Here's what I think this actually signals for business leaders:

𝗧𝗵𝗲 𝗽𝗹𝗮𝘁𝗲𝗮𝘂 𝗶𝘀 𝗿𝗲𝗮𝗹 — 𝗯𝘂𝘁 𝘁𝗲𝗺𝗽𝗼𝗿𝗮𝗿𝘆. The slowdown isn't because AI hit a ceiling. It's because autonomous coding and deep real-world integration are harder than benchmarks suggested. That gap closes. It just takes longer than a research paper assumes.

𝟮𝟬𝟮𝟵 𝗔𝗚𝗜 𝗶𝘀 𝗻𝗼𝘁 𝗮 𝗰𝗼𝗺𝗳𝗼𝗿𝘁𝗮𝗯𝗹𝗲 𝘁𝗶𝗺𝗲𝗹𝗶𝗻𝗲. We are building 3-5 year business strategies right now. If the most credible ex-OpenAI forecaster puts AGI inside that window — even at his "slower" revision — your strategy had better account for it.

𝗧𝗵𝗲 𝗼𝗿𝗶𝗴𝗶𝗻𝗮𝗹 "𝗔𝗜 𝟮𝟬𝟮𝟳" 𝗿𝗲𝗽𝗼𝗿𝘁 𝘀𝘁𝗶𝗹𝗹 𝗺𝗮𝘁𝘁𝗲𝗿𝘀. Kokotajlo still puts AGI by end of 2027 at a non-trivial probability. He's not saying it won't happen soon. He's saying the median shifted two years right. That's a revision, not a retraction.

At TenderB, I've stopped building 5-year roadmaps and started building 18-month ones with explicit "AGI-adjacent" checkpoints. Not because I'm certain it's coming. Because the uncertainty itself demands a different planning posture.

The companies that will survive this aren't the ones who predicted the date correctly. They're the ones who stayed adaptive enough not to need the date.

How are you adjusting your planning horizon for AI uncertainty?

#AI #AGI #FutureOfWork #B2B #Leadership #ArtificialIntelligence #Strategy

The AI "Singularity" isn't a software problem. It’s a physics problem.

And Elon Musk just made the biggest bet in history to solve it. 🚀

Everyone is reacting to the SpaceX x xAI merger. But almost no one is talking about the "Three Walls" that made this move inevitable.

If you think this is just about "better chatbots," you're looking at the wrong map.

The 3 Walls Blocking AI's Future:

💰 The Cost Wall: Training a frontier model now costs $1B+ per run. It’s no longer a "scaling law"—it’s a wealth tax that only three companies on Earth can afford to pay.

⚡ The Energy Wall: Terawatt-scale data centers are crushing local power grids. We are literally running out of "Earth-bound" electricity to feed the models.

📊 The Data Wall: We’ve officially "eaten" the internet. There is no more high-quality human data left to scrape.

Why Space is the "Cheat Code":

This merger isn't a business pivot. It’s a vertical integration of the entire physical stack.

🛰️ Free Cooling: In orbit, you have near-absolute-zero temperatures. Passive cooling for high-density chips isn't just a feature—it’s free.
☀️ Unlimited Energy: Solar power without atmospheric interference means 24/7, high-yield energy that never touches a terrestrial grid.
🌍 Infrastructure Autonomy: Orbital data centers remove the constraints of land, permits, and national borders.

When you combine SpaceX’s launch capacity with xAI’s compute stack and Tesla’s manufacturing, you don't just "manage" the singularity bottleneck.

You bypass it entirely.

The projected IPO range of $300B–$800B isn't about today's revenue.

The market is pricing in the first "Post-Earth" infrastructure layer. The race has shifted from who has the best code to who has the best access to energy and physics.

The next phase of the AI war won’t be won in a data center in Virginia.

It will be won in low Earth orbit. 🌌

Is the future of compute actually "Off-World," or is this the ultimate over-leverage? Let’s debate the physics below. 👇

#SpaceX #xAI #ElonMusk #AIInfrastructure #Singularity #TechTrends2026

AI's Singularity: A Physics Problem

CCO TenderB, turning tenders from chaos into contracts — saving time and giving control back to teams with TenderB

Soms is '𝗻𝗲𝗲' het meest professionele antwoord dat je kunt geven.

We waren in gesprek met een potentiële klant. Een bedrijf met een indrukwekkende historie, waar precisie en veiligheid de kern vormen van alles wat ze doen. Een partij waar ik oprecht enthousiast van werd.

Het product was niet het probleem. 
De klik met het team was er. 
Het vertrouwen in wat wij brengen ook.

Maar ze wilden ons binden aan inkoopvoorwaarden die stammen uit een tijdperk van dictaten in plaats van dialoog. Voorwaarden waarbij de leverancier alle risico's draagt en de opdrachtgever de touwtjes strak in handen houdt via ouderwetse sancties. Geen partnership, maar een keurslijf.

Wat heb ik gedaan? Ik heb het gesprek aangegaan.

Ik heb uitgelegd hoe een moderne samenwerking eruitziet. Gebaseerd op wederkerigheid, vertrouwen en gedeelde verantwoordelijkheid. Want alleen vanuit een gelijkwaardig fundament kun je écht iets opbouwen. De deur bleef dicht. Het was hun weg, of geen weg.

Toen heb ik nee gezegd.

En dat voelde niet als verlies. Het voelde als zelfrespect. 

Dit is precies het gevaar van ouderwets aanbestedingsdenken: het houdt vooruitgang buiten de deur. Bedrijven die inkoop behandelen als een juridisch gevecht in plaats van een strategisch partnership, verliezen uiteindelijk de innovatiekracht en de passie van hun partners. Ze krijgen misschien de laagste prijs, maar nooit de beste samenwerking.

Echte relaties in sales zijn gebouwd op eerlijkheid. Eerlijkheid betekent soms ook: dit past niet bij hoe wij zaken doen. Groeien doe je samen. En als de basisvoorwaarden die groei in de weg staan, dan is nee zeggen geen zwakte maar een bewuste keuze voor kwaliteit. 

Wanneer heb jij voor het laatst nee gezegd tegen een grote kans, omdat de voorwaarden simpelweg niet klopten?

taplio

The Battle

Alex de Ruiter

Steven Tsirakos

Followers

Posts

Engagements

Likes

Performance ScoreAlex de Ruiter's LinkedIn Performance

Ready to grow your LinkedIn presence?

taplio

The Battle

Alex de Ruiter

Steven Tsirakos

Followers

Posts

Engagements

Likes

Performance Score.css-goub6s{font-size:var(--chakra-fontSizes-md);font-weight:500;color:var(--chakra-colors-gray-600);display:block;margin-top:var(--chakra-space-1);}Alex de Ruiter's LinkedIn Performance

Ready to grow your LinkedIn presence?

Performance ScoreAlex de Ruiter's LinkedIn Performance