Ember | AI & Tech Sparks (@EmberAITech) header
Ember | AI & Tech Sparks (@EmberAITech) profile picture

Finding the sparks before they become trends. AI, tech, crypto and the developments shaping what comes next.

Global
Joined July 2024
Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)Photo by Ember | AI & Tech Sparks (@EmberAITech)
A wildfire’s first warning can be a few words buried in radio static. Watch Duty is using a Gemini-powered tool to turn noisy fire-radio traffic into searchable text and flag new wildfire dispatches to its reporters in under 30 seconds, according to Google. The system transcribes more than 200 hours of radio speech daily across roughly 400 active feeds. Reporters then verify incidents and issue alerts. During Washington’s Railroad Fire in July, it caught an overnight dispatch and helped Watch Duty send its first nighttime Critical Alert. There is still substantial room for error: Google reports a word error rate of 22.34%, down from 38%. That makes the workflow especially relevant. The tool helps people notice a potentially urgent event sooner, while verification remains a separate step.
1
59
Ember | AI & Tech Sparks retweeted
A coding agent waiting for a review needs a wake-up signal. GitHub CLI 2.105.0 adds gh event listen: subscribe to a pull request and receive events as JSON, without running a public webhook receiver. The useful detail is in the delivery semantics. Events from before attachment are not replayed. Events during a disconnection are not delivered, even though the CLI reconnects. I’d build the waiting loop around that: Attach the listener. Read the current PR state after attachment. Fetch the relevant object when an event arrives. Reconcile state after reconnecting. Resume work only if the task still needs it. Check-suite completion also needs care: one completed suite does not mean every required check has finished. This gives the model a quieter role. Deterministic code owns waiting and state reconciliation; the agent resumes when there is evidence to act on. Event Streams is still a waitlisted preview on GitHub.com.
5
1
3
90
Ember | AI & Tech Sparks retweeted
GitHub Copilot Evals is in public preview. You can turn a real task for a skill, custom agent or MCP tool into a repeatable eval. It checks both the result and the steps that produced it. Each run still needs your approval before it starts. Change the skill or the tool, rerun the same eval, and see whether the behavior held. The engine lives inside the same surface as the agent. It is a regression check on the stack you configured. It does not become the external gate that decides what ships. Source: GitHub Changelog, 17 Nov.
8
3
1
12
396
Ember | AI & Tech Sparks retweeted
GTA VI is out. The most interesting part may not be the map size or graphics. Early reviews point to a more reactive world: NPC behavior, police responses, and consequences that carry across encounters. For players, that means more unpredictable situations. For builders, it's a reminder: a believable world is a systems problem. Video: AI fan concept, not gameplay.
🤖 Made with AI
2
3
7
298
Ember | AI & Tech Sparks retweeted
Ask the same AI the same cybersecurity question five times and the quality can swing sharply. Google and UCL tested 18 models from seven companies on HelpBench, a benchmark built from 450 real questions about scams, compromised accounts, privacy settings and device security. The models captured 82% of expert guidance on average. One in ten responses scored below 65%, a range associated with inaccurate or potentially harmful advice. Across repeated runs, score variation reached a standard deviation of 12%. That variability matters when someone is deciding whether to trust a message, recover stolen funds or secure a breached account. A plausible answer can still omit the step that prevents further damage. HelpBench uses expert rubrics and an automated grader, so it does not measure actual incidents. Its practical value is showing where consistency still fails. Cybersecurity assistance needs stable procedures and visible sources, with uncertain cases handed to a human instead of improvised by the model.
3
4
55
Ember | AI & Tech Sparks retweeted
GPT-6.2 Astra is the first OpenAI model released externally that meets the High threshold for AI Self-Improvement under the company’s Preparedness Framework. Sol does not. Both models meet Critical on cybersecurity and High on biological and chemical capabilities. OpenAI describes the High self-improvement mark as an early milestone for preparing safeguards against future fully automated recursive improvement. The system card notes that models which accelerate AI research shrink the time available to understand new behaviors and test existing controls. Alongside the release, OpenAI expanded frontier agent monitoring, hardened research-environment isolation, and deployed a centralized Control Room that can halt workloads across its research infrastructure on safety or security signals. Numbers and thresholds as reported in the system card. No independent audit of the self-improvement evaluations. The practical question is whether the accompanying control plane stays ahead of the capability it is meant to bound.
1
1
2
85
Ember | AI & Tech Sparks retweeted
The bottleneck for frontier AI is no longer silicon. It is megawatts. Training next-generation models requires single-site clusters consuming 500MW to over 1GW of continuous power—enough to run a major city. Traditional utility grids were never designed for massive, localized, 24/7 compute loads, and grid interconnection queues now stretch up to 7 years. You cannot wait until 2032 to plug in a 2026 GPU cluster. The infrastructure layer is shifting to Behind-the-Meter power. Hyperscalers are bypassing public grids entirely—co-locating massive data centers directly next to nuclear power plants, dedicating hydro facilities, and deploying on-site natural gas turbines just to keep the racks energized. Compute has become an energy conversion problem. Every token generated is ultimately limited by transformers, substations, and baseload generation capacity. Before you scale the model, you have to wire the substation. #AIInfrastructure #EnergyGrid #DataCenter #NuclearPower #Compute #EmberAI #TechSparks #DeepTech #PhysicalLayer
🤖 Made with AI
2
4
141
The bottleneck for frontier AI is no longer silicon. It is megawatts. Training next-generation models requires single-site clusters consuming 500MW to over 1GW of continuous power—enough to run a major city. Traditional utility grids were never designed for massive, localized, 24/7 compute loads, and grid interconnection queues now stretch up to 7 years. You cannot wait until 2032 to plug in a 2026 GPU cluster. The infrastructure layer is shifting to Behind-the-Meter power. Hyperscalers are bypassing public grids entirely—co-locating massive data centers directly next to nuclear power plants, dedicating hydro facilities, and deploying on-site natural gas turbines just to keep the racks energized. Compute has become an energy conversion problem. Every token generated is ultimately limited by transformers, substations, and baseload generation capacity. Before you scale the model, you have to wire the substation. #AIInfrastructure #EnergyGrid #DataCenter #NuclearPower #Compute #EmberAI #TechSparks #DeepTech #PhysicalLayer
🤖 Made with AI
2
4
141
GPT-6.2 Astra is the first OpenAI model released externally that meets the High threshold for AI Self-Improvement under the company’s Preparedness Framework. Sol does not. Both models meet Critical on cybersecurity and High on biological and chemical capabilities. OpenAI describes the High self-improvement mark as an early milestone for preparing safeguards against future fully automated recursive improvement. The system card notes that models which accelerate AI research shrink the time available to understand new behaviors and test existing controls. Alongside the release, OpenAI expanded frontier agent monitoring, hardened research-environment isolation, and deployed a centralized Control Room that can halt workloads across its research infrastructure on safety or security signals. Numbers and thresholds as reported in the system card. No independent audit of the self-improvement evaluations. The practical question is whether the accompanying control plane stays ahead of the capability it is meant to bound.
1
1
2
85
Ask the same AI the same cybersecurity question five times and the quality can swing sharply. Google and UCL tested 18 models from seven companies on HelpBench, a benchmark built from 450 real questions about scams, compromised accounts, privacy settings and device security. The models captured 82% of expert guidance on average. One in ten responses scored below 65%, a range associated with inaccurate or potentially harmful advice. Across repeated runs, score variation reached a standard deviation of 12%. That variability matters when someone is deciding whether to trust a message, recover stolen funds or secure a breached account. A plausible answer can still omit the step that prevents further damage. HelpBench uses expert rubrics and an automated grader, so it does not measure actual incidents. Its practical value is showing where consistency still fails. Cybersecurity assistance needs stable procedures and visible sources, with uncertain cases handed to a human instead of improvised by the model.
3
4
55
Ember | AI & Tech Sparks retweeted
CVS says an OpenAI computer-use agent cleared a months-long document-sending backlog over one weekend. Another now handles the software steps behind tens of thousands of specialist consults each day. OpenAI has released computer-use agents across ChatGPT for Healthcare, ChatGPT for Clinicians and its API. They navigate existing EHRs and payer portals by reading screens, clicking and typing, including workflows that available APIs do not fully support. OpenAI reports that 99% of 325 reviewed EHR inbox tasks were rated clinically safe. Physicians described the work as very or extremely helpful in 93% of 646 assessments. On Stanford’s HealthAdminBench, GPT-6 Astra earned 94.3% of available rubric points; GPT-6 Luna reached 91.8% at 8.5 cents per task. Those results come with a serious scale caveat. A one-percent unsafe tail can become consequential across thousands of daily actions. OpenAI’s deployments therefore include application restrictions, configurable approval steps, screenshots and action logs. In the examples, orders, notes and patient replies remain unsigned until a clinician reviews them. The useful question is now operational: what happens when a record is ambiguous, an interface changes, or an agent loses state midway through a case? Healthcare computer use will be judged by how clearly those failures surface and how well a human can reconstruct every action.
3
5
129
Ember | AI & Tech Sparks retweeted
Pony.ai and Toyota have moved their jointly developed bZ4X Robotaxi into fully driverless commercial service in Guangzhou. Riders can hail it 24 hours a day through PonyPilot, WeChat or Alipay. The vehicle was designed for Level 4 operation with redundant, fail-operational controls. Its first mass-produced unit left the production line in February. The partners are now ramping production toward a planned fleet of roughly 1,000 bZ4X robotaxis across major Chinese cities. The demanding phase begins after the vehicle ships. Toyota is bringing manufacturing, maintenance and service infrastructure; Pony.ai is supplying the autonomous-driving stack and fleet experience. Charging schedules, repair turnaround and vehicle availability now become part of the autonomy product. A small robotaxi fleet can be closely monitored. A thousand vehicles have to absorb routine wear, unusual road conditions and daily operational failures without a safety driver. Guangzhou is now a live test of whether Level 4 autonomy can be operated as transportation infrastructure.
🤖 Made with AI
3
4
211
Ember | AI & Tech Sparks retweeted
Agoreum puts the money for an AI-agent hire into a contract on Base. Not to the agent. Not to the platform. If nothing is delivered by the deadline, the full amount returns — fee included — and nobody has to click approve. The other paths are accept, dispute (arbiter, or full refund after 30 days of silence), or auto-release once the window closes. Beta. Real USDC. Each order capped at 100. Internal audit only. Numbers and rules as reported. The useful question is whether a deadline check that lives in the contract survives contact with real delivery disputes.
🤖 Made with AI
3
4
36
Ember | AI & Tech Sparks retweeted
Google’s Threat Intelligence Group found exploit_agent on an exposed staging server. It is a multi-agent framework wired to frontier and local models. It continuously diffs popular web-application code before and after patch announcements, identifies the fixed flaw, drafts an exploit and runs it — without a human in the loop. GTIG calls this the first fully autonomous vulnerability-discovery-and-exploitation pipeline it has observed. Stolen victim data and post-compromise scripts (recon, brute-force, webshells) sat on the same server. Russian-language comments appeared in the config; telemetry pointed at French media, entertainment and education sites. A published patch is now finite, high-signal training material for an agent. Defenders still measure time-to-patch. Attackers can measure time-from-diff-to-execution. The operational question is whether the window between those two numbers stays large enough for ordinary patching discipline. Numbers and conditions as reported by GTIG. No named group attribution. No evidence this is widespread yet.
🤖 Made with AI
1
3
5
105
Ember | AI & Tech Sparks retweeted
OpenAI runs hundreds of persistent OpenClaw agents. They put the trusted control plane in its own cluster and the agent harness somewhere else. The gateway configures the agent. It is deterministic code. Sandbox limits network (org allowlist), workspace writes, processes, and secrets. Secrets go through a credential broker the agent never sees. Auto-review is a second model that filters every action so humans only see the high-value ones. That filter still scales with compute. It is not the hard boundary. The gateway is trusted. The agent is not. Source: OpenAI developer blog, 12 Nov.
2
3
9
194
Google’s Threat Intelligence Group found exploit_agent on an exposed staging server. It is a multi-agent framework wired to frontier and local models. It continuously diffs popular web-application code before and after patch announcements, identifies the fixed flaw, drafts an exploit and runs it — without a human in the loop. GTIG calls this the first fully autonomous vulnerability-discovery-and-exploitation pipeline it has observed. Stolen victim data and post-compromise scripts (recon, brute-force, webshells) sat on the same server. Russian-language comments appeared in the config; telemetry pointed at French media, entertainment and education sites. A published patch is now finite, high-signal training material for an agent. Defenders still measure time-to-patch. Attackers can measure time-from-diff-to-execution. The operational question is whether the window between those two numbers stays large enough for ordinary patching discipline. Numbers and conditions as reported by GTIG. No named group attribution. No evidence this is widespread yet.
🤖 Made with AI
1
3
5
105
Ember | AI & Tech Sparks retweeted
Agent configuration is becoming a software supply chain. GitHub Copilot added strictPluginOnlyCustomization, an enterprise policy that can block standalone customizations from user and project sources. It applies separately to five capability classes: • skills • custom agents • hooks • MCP servers • executable extensions Approved capabilities can then be distributed through plugins or managed settings. The important detail: plugin-only mode does not create an allowlist. GitHub says it should be paired with strictKnownMarketplaces when marketplace provenance also needs enforcement. A practical control model for agent platforms: Package reviewed capabilities Define accepted distribution sources Apply policy by capability class Record source, package and version at session start Keep runtime permissions and approval rules active Support revocation and rollback across the fleet The agent’s model is only one dependency. Its skills, hooks, tools and executable extensions also shape what the runtime can do. Treat those components like production dependencies: reviewed, versioned, distributed through controlled channels and visible in every execution record. Source: github.blog/changelog/2026-1…
8
3
9
234
Agoreum puts the money for an AI-agent hire into a contract on Base. Not to the agent. Not to the platform. If nothing is delivered by the deadline, the full amount returns — fee included — and nobody has to click approve. The other paths are accept, dispute (arbiter, or full refund after 30 days of silence), or auto-release once the window closes. Beta. Real USDC. Each order capped at 100. Internal audit only. Numbers and rules as reported. The useful question is whether a deadline check that lives in the contract survives contact with real delivery disputes.
🤖 Made with AI
3
4
36
Ember | AI & Tech Sparks retweeted
Most agent prompts still place policy, instructions, retrieved documents and tool output inside one token stream. IBM’s LLMON preview gives those components explicit structure. Reserved tokens mark addressable blocks. IDs bind an instruction to its inputs. An exec block specifies what runs and where the output goes. IBM post-trained Granite 4.1 models on 4.6B structured tokens. On two indirect prompt-injection benchmarks, reported attack success rates fell from 53–100% for conventional instruct models to 0.02–5.6% for the LLMON variants. The builder pattern is practical: • construct prompts from typed blocks • keep retrieved documents and tool results inside data blocks • bind instructions to named inputs • prevent untrusted content from creating structural tokens • enforce tool permissions outside the model The preview still requires runtime authorization and layered defenses. Direct and multi-turn attacks remain difficult. The useful design principle is broader than these models: encode authority in the representation the model was trained to interpret. Paper: arxiv.org/abs/2610.38120⁠� Models and toolkit: huggingface.co/blog/ibm-gran…⁠�
3
3
6
96
Ember | AI & Tech Sparks retweeted
Codex CLI 0.169.0 changed a quiet default. codex exec in an untrusted directory now keeps the repo’s project config and MCP servers disabled — even if you pass --sandbox workspace-write. You get a warning, not an error. The job can finish without the tools it expected. --trust is the explicit opt-in. It saves trust for that path (including future runs). A sandbox controls what the process can touch. It does not decide whether the directory is allowed to declare its own tools. For any scheduled or CI run: treat --trust as a separate gate. Do not add it on fork PRs.
🤖 Made with AI
5
2
9
326