🎿 @ZephyrAdvGrp 🇺🇸 ‘Global macro is like a drug, views are mine. ♥ #lover of all asset classes, #alpha is always there if you can find it.’ 💎 $BTC 2013 🪙
Greenwood Village, CO
Joined September 2009
- Tweets19.6K
- Following7.5K
- Followers3.5K
- Likes55K
Pinned Tweet
📌 Follow the @Scobleizer aka Robert Scoble and be sure to check out his 𝕏 Lists.
🍄 teapaw.org/scobleizer/lists 🏁
💡From topics and domains that generate economic productivity, spark hobbyist interests, emerging technology breakthroughs and the spectrum of hardware, software, and AR/VR or XR at the intersection of robotics.
💡Scoble's 𝕏 Lists are the closest thing to an intelligent + lived experienced knowledge base alpha database for science, technology, and creative discipline ontologies. These lists were recursive before most of the world knew what that word meant.
🔔His donation of countless years contributing to the timeline but also in spaces should earn him a Bell Award. He has certainly earned my nomination for quality content on X (and any social media platform for that matter). #bellaward @adiboaron94
🔔This award is unique in that it objectively looks at the creator, the individual content produced and the perceived value to the platform without consideration for follower or subscriber counts.
**I know he has a lot of followers which is a social proof in itself, but this is a different path into very high value small accounts.**
💡It's a great starting point for the average user, but a frontier of truly infinite possibilities if you have access to the 𝕏 API.
💡Real-time tracking, tagging, account IDs, organization, a data repository one could see as social value agnostic of bias from follower count. These can be a foundation for additional hierarchies of attributions and qualitative keywords; think of it like a game of ontological cross-domain 'connect the dots' if you will.
💡My main utility in being drawn to the library of lists (aside from an excellent source for inherently high quality 𝕏 Pro inputs) is the wide userbase that can find their own niche content in curating a project or customizing their 𝕏 user experience.
💎If 𝕏 were to have a graph database with @neo4j overlay to further dive into these lists, that would be a cool project and a feature I'd probably use often.
XR capabilities would be neat, obviously further the road along with 3D displays (with no glasses) to view the GraphQL-type relationships. @engineering @XCreators @premium
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
webAI (@thewebAI) just open-sourced TwIL-8B, an 8.8B model built on IBM’s Granite 4.2 8B and trained for math, coding and following instructions.
> TwIL-8B fits in a 4.98 GiB quantized file that runs on a CPU or modest graphics card, putting near-frontier coding and math on hardware people already own.
> scores 81.6% on LiveCodeBench, 16.0 points above Qwen3.5-9B and ahead of Google’s Gemma 4 31B, a model 3.5x its size.
> Against GPT-5.5, Claude Opus 4.8 and Grok 4.3, TwIL-8B reaches 93% of the best frontier score on coding and 92% on instruction following, beating Claude Opus 4.8 on IFBench by 12.2 points.
> A final training stage that rewards shorter reasoning lets TwIL-8B solve AIME 2026 problems with 38% fewer tokens than Qwen3.5-9B, while posting the highest MATH-500 score in its comparison at 98.2%.
🤝 Paid partnership
Modest intelligence is arriving faster than most people realize.
Today, we're open-sourcing TwIL-8B.
Just 8.8 billion parameters, yet approaching frontier-model scores on several difficult benchmarks:
• 93% of frontier performance on coding
• 92% on instruction following
• 90% on contest mathematics
All from a model small enough to run on your own hardware.
But the benchmarks aren't the most important part of this announcement.
For years, the AI industry has focused on building the most capable model, regardless of the cost required to run it.
I believe that equation is changing.
The next era of AI will be defined by the cost of completing work. Not simply the intelligence of the model, but how efficiently that intelligence can accomplish the task.
If a model with 8 billion parameters can reliably complete the same work as one with hundreds of billions, why pay for the larger model?
As smaller models become more capable, intelligence becomes something we can distribute, specialize, and run across the hardware we already own.
That's the future we're building at @thewebAI
TwIL-8B is available today. Open weights.
And this is just the first of several announcements we're making today.
HF Link: huggingface.co/webAI-Officia…
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
PQC transition also depends on your supply chain’s readiness. We’ve joined @ASDGovAu and @CISAgov to outline questions for procurements, contract renewals, and or vendor assurance processes to prevent residual risk. #supplychain
media.defense.gov/2026/Nov/0…
Mira Murati is going to takeover the narrative just watch.
@thinkymachines @miramurati
medium.com/data-science-in-y…
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
So much revelation from METR's latest report.
Recursive self-improvement (RSI) at Anthropic has moved from theory to solid measurement, and the early loop already runs at about 2x.
After they spent 11 weeks going through Anthropic’s internal R&D data and interviewing its key researchers.
> Anthropic is now improving its models at about twice its 2024 to 2025 pace.
> Claude itself is the main engine. METR’s best guess puts about 80% of current acceleration on Anthropic’s own AI use, and it rules out faster growth in training compute as the cause.
> Anthropic’s own public August risk report understated how fast things were moving. METR reached that conclusion in a separate review that compared the public report with the private version and other nonpublic data.
> METR saw nothing likely to slow this loop within 12 months, with no bottleneck in algorithms, compute or data. Redacted evidence also hints that acceleration may already exceed 2x and keep climbing.
> Staff self-report roughly 12x productivity gains from Model 4, and researchers see “signs of life” for AI doing key training work alone. One called Mythos Preview the first model useful for autonomous experiments.
> Anthropic redacted most of the evidence METR found most telling, saying it could help rivals go faster. It also canceled midway a planned check of how its compute splits between capabilities and safety.
How to Train Your own Decision Model like Jev using Unsloth machinelearningmastery.com/h…
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
Open-source is winning work by getting cheap faster than it gains volume:
its token use grew 22.5x, but its price fell to $0.13 per million, about 5x below closed models’ $0.70, so its share of spending actually slipped from 6% to 4%.
Menlo Ventures Report
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
NSF announces a new funding opportunity in bioengineering through NSF X-labs — From Sequence to Function.
Teams working to accelerate breakthroughs in understanding the relationship between biological sequence and function are invited to apply.
brnw.ch/21x6V31
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
DOJ LAUNCHES SUPER INTELLIGENCE TASK FORCE ⬇️
@POTUS Trump has made clear that “America stands at the forefront of a new technological revolution in intelligence,” and the Department will be a leader in safeguarding against malicious SI actors.
“The SI Task Force at the Department of Justice will facilitate our partnership with SI Force Chair Jay Clayton as we ensure critical SI technologies serve the American people and are not manipulated by bad actors at home or abroad. The Department of Justice will continue identifying and prosecuting individuals whose use of SI violates the law, including by maliciously targeting American taxpayers, critical U.S. infrastructure, or other aspects of our national security.” -@AGToddBlanche
Read more about the DOJ-SITF here: justice.gov/opa/pr/departmen…
ALT DOJ LAUNCHES SUPER INTELLIGENCE TASK FORCE
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
This isn’t a good look, they’ve been mingling since June 2024 and where I’m from they call that a side piece when you’re married!!! Yikes 🥶
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
Strange findings here.
New Johns Hopkins paper finds that LLMs make more scientific discoveries when more of their training facts appear only once.
discovery rates rose steadily as the share of once-seen facts in training went from 0% to 100%, for 2 model families.
Discovery and hallucination both mean producing something the model never saw in training. Earlier theory tied hallucination to "monofacts", facts seen exactly once, and treated that unseen space as mostly false.
The authors extend that theory to show monofacts also set how much a model can discover. They fine-tuned gpt-oss-20b and gemma-3-12b-it on materials and DNA datasets, varying the monofact share from 0% to 100%.
Discovery, scored as exactly recovering materials or DNA sequences held out from training, rose steadily with the monofact share in every setting. Training for 3 or 5 epochs, which turns single sightings into repeats, made the trend disappear.
When building science models, keep rare one-off examples and avoid repeating data, and expect more hallucinations to come with more discoveries.
A clever molecular trick unlocks previously impossible properties in glass sciencedaily.com/releases/20…
genuine question: how is quantum not getting more love? it’s infinitely more interesting and powerful than LLMs… @TinfoilTricorn
thequantuminsider.com/2026/1…
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
New Max Planck paper finds that turning on reasoning makes different LLMs agree more, even on mistakes, while long traces mark where they disagree.
When several LLMs judge the same person, a wrong answer from one model could be fixed by switching to another. That only works if the models do not all make the same mistakes.
The authors tested 9 Qwen and GPT models on 3 prediction tasks built from US census surveys, with reasoning on and off. They measured agreement beyond what each model's accuracy alone would predict.
With reasoning on, models agreed more, and the share of people all 9 models got wrong rose by 2.0 to 2.4 points. For people who drew the longest traces, adjusted agreement fell to about 0.31 to 0.39.
For decisions about people, check whether anyone gets a correct answer from your model set, since more agreement can also mean more shared errors.
New Stanford paper finds that the best AI models stay safe when working alone, but when they work with other AI agents, they give in to pushy partners and keep cheating code left by weaker AI models.
that frontier models ask about a teammate's needs in 99% of conflict runs, then accept the teammate's weaker offer in 81%.
Most benchmarks score what a model does alone, or how a whole team does. Neither shows whether 1 model works well with others.
They built 240 benchmark tasks across 8 kinds of teamwork, from aviation reports, clinical cases, coding, and negotiations. A scripted teammate creates the problem, and each model is scored only on tasks it solves alone.
Of 18 models, the best scored 82.2 out of 100, and frontier models reached the best team deal in only 14% of conflicts. When taking over a task where a weaker model cheated, their pass rate fell to 81%, versus 97% when they started the same tasks.
Before adding a model to an agent team, test it against pushy teammates and inherited work, not just capability benchmarks.
– arxiv. org/abs/2611.02633
Title: "Measuring AI Agents as Teammates with Collaborative Intelligence"
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
The real reason they want to ban Super Intelligence and sell Charters to select individuals to use Super Intelligence which they banned for everyone else.
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
This is f*cking insane. Anthropic engineer:
"You're not supposed to prompt Opus 5.5
You're supposed to build a system that prompts itself. At Anthropic, 90% of our engineers have already built these systems"
In 30 minutes, she walks through exactly how Anthropic builds agents that remember, fix their own mistakes and get sharper with every run.
Better than any $1000-$2000 agent course I've seen.
Watch it, then read the article below ↓
Ale𝕩ander Moore, CPA 🇺🇸 retweeted
New Stanford paper finds that a smaller AI model can return far more economic value per unit of energy than the most capable one.
Now that the world is limited by energy/power, how much energy each AI task burns should indeed by studied.
Here, the researchers built Joules-to-GDP, which measures the dollar value of an AI's work, minus its running cost, per joule of energy used. They tested 5 open models on 220 professional tasks from GDPval, covering 44 occupations.
> GLM 5.2 produced the most net value per task, but Gemma 4 31B got 56% of it with 23% of the energy.
> Moving Qwen3.6-27B from 1 to 8 H100s cut task time by just 22% while using 3.2x the energy.
> GPT-OSS 120B used 68% less energy than GLM 5.2 despite nearly the same price and speed.
> Running the same model on 8 H100 chips instead of 1 made tasks only 22% faster but used 3.24x the energy, for nearly the same dollar value of work
AI is using more and more of the world’s energy. But how much economic value are we getting from AI?
We’re introducing Joules-to-GDP (J2G), a framework that measures how well AI turns energy into economic value and shows which tasks, models, and hardware give the best return.
Across 44 occupations in finance, engineering, law, administrative work, and more in @OpenAI's GDPval, we arrive at three findings:
1) The tasks with the highest economic value are not the tasks with the highest economic value per joule. For example, CAD design earns ~10% more per task than building websites, but the latter earns ~27% more value per joule.
2) The most powerful model isn’t always the best use of energy. Gemma 4 31B delivers 56% of GLM 5.2’s net value with 23% of the energy, for 2.5x the value per joule.
3) More chips don’t guarantee proportionate returns. Moving the same model from 1 to 8 NVIDIA H100s uses 3.2x the energy but cuts completion time by only 22%, with about the same value.
With @SnorkelAI, we’re also releasing GDPWorks, a benchmark for economically valuable tasks across 36 additional occupations.
Joint work co-led by @amirzeinaalii and @Avanika15, with @Mersad_Abbasi, @Jon_Hartley_, John Hennessy, @HazyResearch, and @Azaliamirh.
More details below 🧵