Illustration: loose strands of light, what AI can do next, draw together at one point, design and build, into a single braid: what holds up. A few strands peel away in coral: what didn’t.
I’m Jason. I research what AI can do next, then build the systems and experiments to find out what holds up.
I built a harness that compares both the response and the resulting state against a real system: a pinned local Gitea service (an open-source Git server).
A mock can agree on the reply and still leave a different state behind.
Attribution minus the strongest cheap baseline (sufficiency, percentage points)
Attribution minus the strongest cheap baseline, by taskDot-and-whisker chart, SAE-feature basis and raw-neuron control. On all seven single-token tasks, every estimate sits within 6.0 points of zero. On IOI, where the signal is distributed, the gap is +30.1 points (95% interval +16.0 to +45.8) in the SAE-feature basis and +17.2 points (95% interval +8.7 to +28.1) in the raw-neuron control.0+20+40CapitalsCountry to languagePast tenseComparativePluralAntonymsSuccessorIOIDistributedanswer differs hereSAE featuresRaw-neuron control+30.1+17.2
IOI, distributed
SAE features
+30.1 [+16.0, +45.8]
Raw-neuron control
+17.2 [+8.7, +28.1]
Prompt pairs
18
Pick any task on the chart.
Attribution minus the strongest cheap baseline (sufficiency, percentage points). Gemma-2-2B, layer 7, top 64 units, paired-bootstrap 95% intervals.
Task
SAE-feature basis
Raw-neuron control
Prompt pairs
Capitals
+6.0 (95% interval +2.3 to +10.1)
+3.0 (95% interval +1.7 to +4.3)
20
Country to language
+0.3 (95% interval −2.8 to +3.8)
+2.5 (95% interval +1.6 to +3.4)
20
Past tense
+2.1 (95% interval +0.3 to +4.2)
+2.5 (95% interval −0.2 to +5.3)
30
Comparative
−1.3 (95% interval −3.9 to +0.6)
0.0 (95% interval −1.8 to +1.5)
24
Plural
−2.1 (95% interval −5.8 to +1.9)
+1.0 (95% interval −1.5 to +3.4)
16
Antonyms
−5.8 (95% interval −14.6 to +0.7)
+1.2 (95% interval −1.6 to +3.9)
24
Successor
−0.5 (95% interval −5.8 to +6.6)
−5.0 (95% interval −11.9 to +2.6)
11
IOI (indirect object identification, distributed)
+30.1 (95% interval +16.0 to +45.8)
+17.2 (95% interval +8.7 to +28.1)
18
Gemma-2-2B, layer 7: attribution minus the strongest cheap baseline, for SAE features and a raw-neuron control, with paired-bootstrap 95% intervals. Attribution pulls clear only on IOI (indirect-object identification), the distributed circuit.
Design choice
I grade every method against the strongest cheap baseline I can build, and measure the known answer instead of assuming it.
create_guarded_issueAgenthas resultToolno new writeExecution recordSQLitecase-next-resultappliedreply lostget_operation_statusreturn_stored_resultIssues created1Still 1. No second issue.
Step through what happens next.
Result recorded
The work is done.
The agent calls create_guarded_issue. SQLite commits the issue and its operation record in one transaction.
Caller uncertain
The reply is lost.
The server exits after the commit, before the reply arrives. The agent cannot tell whether the write happened.
Stored result returned
Recover the result.
After a restart, the agent calls get_operation_status with the same operation_id. The answer is return_stored_result, and nothing runs again.
The tool finished the work, and the agent never heard back. A blind retry would create the issue twice. I built a reference that asks the execution record what happened before doing anything again.
A lost reply means unknown, not failed. Ask the system of record before retrying anything with side effects.
New in October: a system an agent handed over, silent scenes directed through timing, and a studio on a busy street. Each card runs a small version of the real mechanic. The full builds open in your browser.
clock
Every fourth tick fires; delivery lands two ticks later.
After the Agent
Take over a light installation an agent set up, and keep it working.
I want to understand what a new capability can do. Building with it is how I find out.
Jason Lovell · Austin, Texas
Each wave, the same method.
From 2004
Mobile
I started in mobile, when the consumer-tech frontier moved one new category at a time: operators first (O2, T-Mobile, EE and TalkTalk), then product and portfolio leadership at the device maker KAZAM.
Connected devices and immersive computing
IoT, wearables, tablets, connected home, smart audio, then VR and AR, including senior product management for VR, wearables and SmartThings at Samsung. Leadership teams across sectors, UK first, then the US.
2016 to 2021
Founding Captivate
I founded Captivate, an applied-innovation consultancy for XR, AI and emerging technology, and ran it for five years. Clients included Synthesia, Gucci, NBC Universal and Lloyd's of London; for Synthesia I was VP of global partnerships in 2018 and 2019. That is where AI caught me: early generative models for synthetic media, alongside immersive work. In parallel, from 2017 to 2018, I led brand partnerships across EMEA for the VR company Jaunt.
2017
Pulled into AI
AlphaGo, OpenAI's Dota work and AlphaGo Zero. I have read the field closely since.
2019 to 2022
Emerging technology and XR strategy, PwC UK
I led strategy and client work on XR and spatial computing where it met AI and data, with a specialist team building training simulations and spatial analysis.
2022 to 2024
AI and emerging technology, PwC US
An international assignment to PwC US in Austin. It began with immersive technology and grew into AI and generative AI: commercialising both for clients and co-developing solutions with technology alliance partners.
2024 to 2025
Technology business partner, PwC UK
Back in the UK, the link between PwC's central technology team and its consulting teams: shaping AI and generative-AI propositions for clients, and bringing together the people to deliver them.
Sep 2025 to now
Frontier AI strategy and enterprise innovation, PwC US
A move to PwC US. First, leading agentic AI delivery for clients; now, with PwC Intelligent Enterprise, frontier research, product direction and guiding multidisciplinary teams: taking an idea from an early research question to a tested prototype and a credible path to scale.
At PwC Intelligent Enterprise my work is frontier research, product direction and guiding multidisciplinary teams, helping organisations rethink how work gets done when AI becomes part of the operating model. I stay hands-on, building the early prototypes that test an idea. The studies and builds here are my own, independent of that role.
The pattern I keep seeing
The arc tends to repeat: loud enthusiasm, then a small number of teams quietly figuring out what is real and shipping it.
How I build
I build everything here myself, end to end, with Claude Code and Codex. I write the question and the test first, read the runs, and keep the results that go against me. Several studies here end negative or inconclusive, and they say so.
What I care about now
The space between an emerging capability and a redesigned enterprise. That is where evidence, judgment and thoughtful implementation matter just as much as the technology.
Releases12 citable releases
Written down, so you can check it.
Specifications, a white paper and research software, each published with the reasoning beside the code, so someone else can reproduce it or build on it.
A harness that counts an AI-generated Metal kernel as faster only after it passes a float64 correctness oracle and a bootstrap confidence-interval timing rule.
A from-scratch lab for one question: if two interactions flatten to the same final chat transcript but require different next actions, what can a transcript-only model know?
Deliberately vulnerable and deliberately clean agent release fixtures, with expected findings and Agent Assurance Case templates, for testing agent-asset assurance tools.