LDraw Nova, Carlos Antelo's open-source tool, had Claude Opus 5.5 design a 2,175-piece LEGO garden by writing a Python program that emits the CAD file. Nova checks parts for collisions but not stability, so whether any of its designs would stand up in real bricks is still untested.
Reality
- Evidence35
- Adoption
- Insufficient
- Hype gap+25
- Incentives
- Insufficient
- Confidence40
Ataraxos, a Stratego AI from Carnegie Mellon, NYU, Stanford and MIT, beat Pim Niemeijer 15-1-4 after a training run that cost under $8,000. The savings came from a custom simulator and sample efficiency, so the price holds only for problems that simulate as cheaply.
Reality
- Evidence62
- Adoption
- Insufficient
- Hype gap+25
- Incentives40
- Confidence60
Ken Griffin is giving Carnegie Mellon $3 billion, a record by the university's count, with $2 billion for a 35-acre campus in Miami's Wynwood. For employers it promises a new supply of technical graduates in South Florida, on a timeline measured in years and at a modest scale.
Perspective Coverage
3 publishers
- Builder
- Builder 23%
- Operator
- Operator 37%
- Investor
- Investor 40%
Reality
- Evidence66
- Adoption
- Insufficient
- Hype gap+18
- Incentives70
- Confidence68
Ken Griffin gave Carnegie Mellon a record $3 billion, $2 billion of it to build a Miami campus that plans to enroll students in 2028. Whether it becomes a research base depends on faculty hiring and industry sponsors arriving on a schedule the money cannot set.
Reality
- Evidence60
- Adoption
- Insufficient
- Hype gap+35
- Incentives72
- Confidence62
Omni Calculator found 54% of managers run at least 70% of their writing past AI, against 30% of experienced individual contributors. The studies tying that habit to weaker thinking did not study working managers. The case for retraining leaders is still an inference.
Reality
- Evidence45
- Adoption55
- Hype gap+35
- Incentives
- Insufficient
- Confidence40
Ataraxos, an AI from MIT, Carnegie Mellon, NYU and Stanford researchers, beat top-ranked human Stratego players by a large margin, MIT reports. The team says it was far cheaper to train than rival systems and also beat top players at other strategy games.
Reality
- Evidence74
- Adoption
- Insufficient
- Hype gap+22
- Incentives70
- Confidence66
Carnegie Mellon senior Nivedita Shyamsundar tracked colour changes in iPhone video frame by frame to show metal nanoparticles form from merging clusters. Her program lets a lab time the fast early stages of such reactions with a camera it already owns.
Reality
- Evidence40
- Adoption8
- Hype gap+20
- Incentives55
- Confidence45
The pilots at Genentech, Janelia and QuEra are the test of whether one agent interface can retire per-device drivers. Its best number, 99.3% autonomous laser relock, comes from the easiest failure mode in the set.
Perspective Coverage
4 publishers
- Builder
- Builder 41%
- Operator
- Operator 38%
- Investor
- Investor 21%
Reality
- Evidence45
- Adoption18
- Hype gap+25
- Incentives65
- Confidence55
Carnegie Mellon's Synthetic Hospital benchmark found agent wrappers often lowered scores for 10 models reading 1,268 synthetic multi-visit patient charts. The models found facts in the charts but struggled to combine them, so a scaffold that adds retrieval steps may be working on the wrong weakness.
Reality
- Evidence50
- Adoption
- Insufficient
- Hype gap+8
- Incentives
- Insufficient
- Confidence45
Hazelwood Green's soil carried BTEX for decades before its cleanup. A microbiologist two miles away is sequencing cores from differently contaminated parts of the site to test whether the resident bacteria kept the genes to break those compounds down.
Reality
- Evidence48
- Adoption15
- Hype gap+30
- Incentives62
- Confidence55
An Entrepreneur contributor who runs a data and AI company says he still believes enterprises need clean data and real governance. He now puts a decision about how staff think alongside AI ahead of it.
Reality
- Evidence25
- Adoption
- Insufficient
- Hype gap+40
- Incentives68
- Confidence42
Two teenagers were arrested in early September after the robotaxi pulled over and alerted emergency services. Waymo described the policy in an email, and left out the confidence threshold and the human review step.
Reality
- Evidence68
- Adoption55
- Hype gap+12
- Incentives58
- Confidence64
A Science paper measured genome folding and gene activity in the same single cells from postmortem prefrontal cortex, and its authors place chromatin architecture alongside amyloid and tau in the disease's molecular pathology.
Reality
- Evidence58
- Adoption
- Insufficient
- Hype gap+25
- Incentives70
- Confidence55
Anthropic's December 2024 guidance and three agent benchmarks converge on an awkward result, because the workloads whose steps cannot be enumerated in advance are also the ones where measured agent completion is lowest.
Reality
- Evidence52
- Adoption28
- Hype gap+18
- Incentives35
- Confidence45
Theory predicted an in-plane anomalous Hall effect and no experiment had shown one, until a few-layer tantalum iridium telluride flake borrowed magnetism from the layer beneath it and produced the signal.
Reality
- Evidence34
- Adoption
- Insufficient
- Hype gap+42
- Incentives74
- Confidence52
The Carnegie Mellon team has released a 13.5-billion-year hydrodynamic run that ends at the present day, which lets JWST's crowd of early massive black holes be tested against a model that also has to produce a recognizable today.
Reality
- Evidence55
- Adoption30
- Hype gap+20
- Incentives70
- Confidence45
PostgreSQL still reported 0.921 correlation and a 48 kB index at that point, so the damage only showed up in the execution plan, where recheck threw away 3.8 million rows to return 111,112.
Reality
- Evidence63
- Adoption
- Insufficient
- Hype gap+20
- Incentives74
- Confidence55
Sprint records come out of motors and simulation. The dexterity events at the World Humanoid Robot Games test the one thing neither of those solves, and a very large field now has to chase them.
Reality
- Evidence48
- Adoption34
- Hype gap+32
- Incentives71
- Confidence55
The lab named tamper resistance an open question 40 days after Inkling's weights shipped. The award is platform compute, priced by the grantor and spendable only on the approved study.
Reality
- Evidence58
- Adoption24
- Hype gap+28
- Incentives79
- Confidence52
NASA's Nancy Grace Roman telescope can launch as soon as August 30 with Hubble-class resolution across a field about 100 times larger. Those pixels have to go somewhere.
Reality
- Evidence42
- Adoption
- Insufficient
- Hype gap+32
- Incentives58
- Confidence38