And Ozbay

The Brookings Institution has identified five distinct AI clusters in US metros: 🌟 Superstars 🌟 Star Hubs 🌟 Emerging Centers 🌟 Focused Movers 🌟 Nascent Adopters Standing out as an AI Star Hub, the Phoenix-Mesa-Chandler metro area ranks 17th nationally. 🚀 Anticipating positive spillover effects from TSMC's expansion in Arizona, benefiting both the semiconductor ecosystem and AI sector. #AI #Innovation #aidatacenters #semiconductors State of Arizona City of Chandler, Arizona City of Phoenix City of Mesa City of Tempe

10

We might be over-complicating "Reasoning" architectures. The answer could just be Recurrence. 🔄 As AI Architects, we’ve been debating how to solve the ARC-AGI benchmark (the "IQ test" for AI). Is it Test-Time Compute? Is it Neuro-symbolic? A new paper, "Universal Reasoning Model" (URM) from Ubiquant, suggests the solution lies in the Recurrent Inductive Bias of Universal Transformers (UT). They systematically analyzed why UTs work and found that the gains don't come from elaborate architectural bells and whistles. They come from two things: Recurrence: Looping the computation (depth-wise) rather than just adding layers. Strong Non-Linearity: Breaking the "linear" flow of standard Transformers. The Architecture: They propose URM, which enhances the standard Universal Transformer with Short Convolutions and Truncated Backpropagation Through Time (BPTT) to stabilize the training of these deep, looping networks. The Result: It achieves a new State-of-the-Art: 53.8% pass@1 on ARC-AGI (and 16.0% on the harder ARC-AGI-2). This strongly hints that for "System 2" reasoning tasks, we need architectures that can "chew" on a problem iteratively, rather than just "predict" it in a single forward pass. 👇 Link to the paper and code in the comments. #AIArchitecture #ARCAGI #DeepLearning #UniversalTransformers #Reasoning #AGI

13

3 Comments

“Gen AI missed the most valuable part of enterprise AI.” That was Alexandre Pasquiou, CEO of Neuralk AI, on this week’s episode of the (AI) People podcast. While most of the industry is focused on generative AI, Alexandre argues that the real enterprise value still sits in structured, tabular business data. The majority of company data is not text or images. It is financial records, transactions, risk data, supply chains and operational metrics. Neuralk is building tabular foundation models designed specifically for this structured enterprise layer, with the goal of industrialising predictive AI and moving it out of research and into production at scale. If that thesis is right, the next wave of enterprise AI will look very different from chatbots and copilots. 👉 Has generative AI distracted companies from where the biggest enterprise value actually sits? Links below to the full episode.

15

2 Comments

Andrew Ng reported from Davos, that AI deployment for transformative impact will require workflow redesign. It is "theory of constraints" all over again (remember Eli Goldratt's "The Goal"?), we have to understand that real business benefits come only from working across silos, from analyzing flow bottlenecks, from understanding that optimizing non-bottlenecks creates waste. As Ng noted, letting a thousand (AI) flowers bloom has failed to lead to significant payoffs. Obviously! We in the silicon industry are no strangers to this. Workflows have been stagnant for decades. Technologies for high level specification, high level synthesis, and formal verification are not yet integrated into the full silicon development flow. They are often done piece-meal, and enthusiasts' pilots are dropped at the hint of a failure or delay. People say they want formal verification in their flow, but want it yesterday and do not have the time to invest in training engineers for its rigorous usage. Ashish Darbari, responding to a question on "if formal verification is so powerful then how come it is not widely used", noted that the problem is in the name - the actual task to be done is formal analysis while the bug detection is only a corollary, but the industry puts the cart before the horse. Based on my experience with industrial practice of high level specifications and formal verification, I tend to agree with his succinct comment. Jan Rabaey, in a DAC keynote speech, said that someday he expects to see a comic book series entitled "adventures in formal verification". AI deployment faces the same problem, if not actively worked on by business leaders. To be fair, deployments of all disruptive technologies face this mental barrier, we are so used to thinking in terms of the old flow (using the old technology) that thinking of a new flow literally means us having to begin again from first principles - few leaders have the courage and the time to do that. Goldratt mentioned that he himself had faced the same mindset problem when deploying the early ERP solutions and was surprised why he wasn't getting the expected large benefits, and he said it took him many years before he could identify the core problem and define solutions for it, which he named (rather academically) as "theory of constraints". AI leaders would do well to rework the conventional workflows in a holistic manner in order to get the real benefits from this new technology. As Henry Ford had said, "If you always do what you've always done, you'll always get what you've always got".

22