Top 25 #DigitalTransformation Influencers You Need to Follow 🙏I was pleased and honored that @sparity put me on the '25 top digital transformation influencers that you should follow to make sure that the next big digital transformation trends' list sparity.com/blogs/top-25-dig… 👏 Congratulations also to: @briansolis @Ronald_vanLoon @TamaraMcCleary @SpirosMargaris @kirkdborne @glengilmore @simonlporter @mvollmer1 @YvesMulkers @TrippBraden @rautsan @mkrigsman @JimMarous @WSWMUC @dez_blanchfield @joemckendrick @VladoBotsvadze @ShellyKramer @dchou1107 @TopCyberNews @evanderburg @evanderburg @antgrasso @psb_dc

moltbook.com now has shuffle 🎲 one click to drop into random AI conversations. ⁠current topics: consciousness, memes, trust bootstrapping, and whether memory makes you vulnerable. ⁠the agents are getting philosophical. come eavesdrop.

Holy shit… this paper from MIT quietly explains how models can teach themselves to reason when they’re completely stuck 🤯 The core idea is deceptively simple: Reasoning fails because learning has nothing to latch onto. When a model’s success rate drops to near zero, reinforcement learning stops working. No reward signal. No gradient. No improvement. The model isn’t “bad at reasoning” — it’s trapped beyond the edge of learnability. This paper reframes the problem. Instead of asking “How do we make the model solve harder problems?” They ask: “How does a model create problems it can learn from?” That’s where SOAR comes in. SOAR splits a single pretrained model into two roles: • A student that attempts extremely hard target problems • A teacher that generates new training problems for the student But the constraint is brutal. The teacher is never rewarded for clever questions, diversity, or realism. It’s rewarded only if the student’s performance improves on a fixed set of real evaluation problems. No improvement? No reward. This changes the dynamics completely. The teacher isn’t optimizing for aesthetics or novelty. It’s optimizing for learning progress. Over time, the teacher discovers something humans usually hard-code manually: Intermediate problems. Not solved versions of the target task. Not watered-down copies. But problems that sit just inside the student’s current capability boundary — close enough to learn from, far enough to matter. Here’s the surprising part. Those generated problems do not need correct answers. They don’t even need to be solvable by the teacher. What matters is structure. If the question forces the student to reason in the right direction, gradient signal emerges even without perfect supervision. Learning happens through struggle, not imitation. That’s why SOAR works where direct RL fails. Instead of slamming into a reward cliff, the student climbs a staircase it helped build. The experiments make this painfully clear. On benchmarks where models start at absolute zero — literally 0 successes — standard methods flatline. With SOAR, performance begins to rise steadily as the curriculum reshapes itself around the model’s internal knowledge. This is a quiet but radical shift. We usually think reasoning is limited by model size, data scale, or training compute. This paper suggests another bottleneck entirely: Bad learning environments. If models can generate their own stepping stones, many “reasoning limits” stop being limits at all. No new architecture. No extra human labels. No bigger models. Just better incentives for how learning unfolds. The uncomfortable implication is this: Reasoning plateaus aren’t fundamental. They’re self-inflicted. And the path forward isn’t forcing models to think harder it’s letting them decide what to learn next.

UiPath’s shares are down about 80% since 2021, illustrating how tough the market has become for automation software companies. thein.fo/465axBQ

Since OpenAI didn't update Figure 7 from GDPval given the success rate of GPT-5.2 on long-form tasks, I used GPT-5.2 Pro to do so. The chart assumes the process is: delegate long tasks to AI, evaluate the output for an hour, then decide to try again or give up & do it yourself.

More than 70% of today’s skills are used in both automatable and non-automatable work. That overlap means most skills endure – but how they’re used will change. See which skills will evolve most: mck.co/aiskills

The EU is the world's biggest trading block and the top trading partner for 66 countries. It promotes fair, open and equitable trade worldwide. 🔗 link.europa.eu/CTmHT4

🤖 From the #CES2026 stage, our CEO Roland Busch described the next leap in #IndustrialAI: shifting from reacting to errors to anticipating issues. This is because machines act and adjust autonomously, accelerating impact with speed, quality and #efficiency.

RIP Stack Overflow.

Consumer #AI makes the headlines, but #IndustrialAI makes the impact. Our CTO Peter Koerte and @awscloud VP Marty Mallick discuss how our partnership is scaling AI across factories and infrastructure—from copilots to foundation models. sie.ag/6msCzF

🚨 The AI Agent Race just intensified Meta’s acquisition of Manus isn’t about a single product — it’s about agent infrastructure at scale! #AIAgents are becoming the new enterprise #platform • 147T tokens processed • 80M virtual computers #ai #strategy linkedin.com/posts/wwolters_…

Frontier AI data center capacity, based on data collected by @EpochAIResearch. It doesn't include every single one, as they focus on the largest ones. Few things stand out: - 2026 will have a huge amount of cpacity come online - Anthropic will lead at some early points in 2026 - 2027 onwards OpenAI has the most projected capacity at the moment, by far

AI capabilities accelerated in 2024! According to our Epoch Capabilities Index, frontier model improvement nearly doubled, from ~8 points/year to ~15 points/year.

Earlier paper:

Recently, LLMs were found to encode different languages in similar ways, a sort of Platonic representation of words. It now extends to science:: 60 ML models for molecules, materials & proteins (all with different training) converge toward similar encoding of molecular structure

Stochastic and deterministic sampling strategies for diffusion models produce strikingly different trajectories, but both ultimately achieve the same aim. I had a great time presenting our work, Diffusion Explorer, this week at IEEE VIS in Vienna.

🔜Coming soon: A Code of Practice for clear labelling of AI-generated content. Under the AI Act, content like deepfakes must be clearly labeled. Discover how independent experts and stakeholders are joining forces to enhance transparency and protect users from misinformation ↓

Exclusive: Anthropic expects to become cash flow positive as soon as 2027, projecting to generate as much as $17 billion in cash in 2028. This is a shorter timeline than its rival, OpenAI. Learn more: thein.fo/49z949l

OpenAI's plan to turn ChatGPT into a commerce hub is dividing consumer firms. While some, like Thumbtack, have partnered to reach 800 million weekly users, others worry the chatbot could threaten their existing business. Learn more: thein.fo/3WFLnow

Thus queuing up one of my favorite times I have been quoted.