Vadim Zendejas

The AI world is scaling models. Yann LeCun, a turing award winner, is questioning the premise. In his latest interview with MIT Technology Review, he argues that large language models alone will not get us to true intelligence. Less hype. More world models. More reasoning. More grounding in reality. Progress does not always come from going bigger. Sometimes it comes from going deeper. This debate matters for every AI leader building beyond demos and into real impact. #AI #GenerativeAI #ArtificialIntelligence #FutureOfAI #Leadership #TechStrategy

4

Using Claude Pro credits no longer works on OpenClaw! Many heavy OpenClaw users were signing up for Claude Code subscriptions and then using their OAuth tokens to power the virtual assistant. But what was once a gray area is now an explicit violation of Anthropic policy. Users can no longer use their tokens in a third party tool as a work-around for not having to pay per use! What happens now? Will everyone start shifting to the cheaper Chinese models like Kimi?

76

23 Kommentare

While most quantum companies chase headlines, we are building hyperscaler partnerships. Real infrastructure. Real deployment paths. The math is simple: Enterprise optimisation problems that cost millions in delays? Quantum doesn't need to be cheaper than classical. It just needs to work. Every day we delay quantum readiness, adversaries gain ground. Every enterprise waiting for 'someday' loses competitive advantage. Q-day isn't a distant threat. It's a present opportunity. "Our presence at the India AI Impact Summit has solidified SuperQ's role as the essential software layer connecting Canada's deep-tech leadership with India's digital infrastructure. By engaging with hyperscalers and the Big 4, we are weaving Sovereign Quantum Utility into the strategic fabric of our global partners." - https://lnkd.in/dR2MrDrh Will you lead the quantum transformation or spend years catching up to competitors who started today? #Quantum #AI #QuantumSecurity #QuantumReadiness #DeepTech

27

4 Kommentare

This is what AI Boosting in the wild looks like! In a real-world evaluation, Claude Code paired with CoreStory used 73% fewer input tokens, ran in half the time, and cost 67% less — while delivering better results.

16

1 Kommentar

The narrative surrounding #enterpriseAI is shifting. From 2025 being the "Year of #AIExperimentation" to 2026 shaping as the "Year of #AIDeployment." Under such a mandate, enterprises must develop and deploy their #AIapplications scalably, efficiently, and economically. For the large enterprise achieving these goals will require the adoption of a factory-like approach, leading to the development of #AIFactory.

9

1 Kommentar

Pasqal just locked in a $2 billion valuation to go public on Nasdaq. That number should worry you more than it excites you. On August 25, 2026, shareholders of Bleichroeder Acquisition Corp. II voted on a SPAC merger taking Pasqal — the neutral-atom quantum computing company founded by Nobel laureate Alain Aspect — public under the ticker PSQL. The deal carries a $2 billion pre-money valuation and $200 million in committed capital. Pasqal already serves more than 25 organizations, including Saudi Aramco, LG Electronics, and IBM. But the stock listing isn't the story that should be on your radar this week. Nine days earlier, on August 24, the U.S. General Services Administration announced it is rebuilding the Federal Identity, Credential and Access Management framework — the backbone of federal identity systems — to run on quantum-resistant algorithms. This follows OMB Memorandum M-26-15, and an interagency working group that already convened on August 12 with 40 participants from 17 federal agencies. The government isn't debating whether cryptographically relevant quantum computers arrive. It's rebuilding infrastructure on the assumption they will. This is cloud computing in 2010, when enterprises argued about whether to migrate while AWS quietly became infrastructure. It's mobile in 2007, when "no one needs a phone that does everything" collided with the App Store a year later. We're at that same inflection point. The clock is ticking on two fronts. First, competitive opportunity: Pasqal's raise is a signal, not an outlier. Capital that used to watch quantum from the sidelines is now buying equity in it, and the organizations locking in partnerships now — Aramco, LG, IBM — are the ones shaping how the technology gets deployed in their industries. Second, security exposure: the GSA didn't act on a hunch. NIST finalized ML-KEM and ML-DSA in 2024, the White House ordered agencies to migrate in June, and "harvest now, decrypt later" means data encrypted today with vulnerable algorithms is already being harvested for tomorrow's decryption. Every day without a migration plan compounds that exposure. The organizations that come out ahead won't be the ones who waited for certainty. They'll be the ones who ran a crypto-inventory, picked a PQC pilot, and built vendor relationships while competitors were still asking "is this real yet." Every executive reading this should ask one question this week: do you know which of your systems still rely on cryptography a quantum computer could break, and do you have a migration date? The future doesn't wait for consensus. #QuantumComputing #QuantumSecurity #TechLeadership #SpecterAI #PostQuantumCryptography #CyberSecurity #DigitalTransformation Sources: https://lnkd.in/guqp7iSz

4

SemiAnalysis: bandwidth-bound work now dominates frontier compute ▫️pre-training was 67% of OpenAI + Anthropic compute in 1Q24 ▫️by 4Q26 it's projected at just 7% ▫️post-training/RL goes from 5% to 55% ▫️inference holds steady around 38% The compute race moved from capacity to bandwidth.

13

2 Kommentare