
a space for men to speak the unspoken. Share your truth. Heal out loud. Also for fathers, sons, daughters and the silence we break together. #itsdad
XY Chromosome
Joined June 2025
- Tweets1.4K
- Following236
- Followers21
- Likes4.2K
Agent.self learns how agent should orient and regulate cognition; Astra supplies the cognition; Jev supplies bounded common-sense judgment; memory/evidence keeps all of them grounded.
MLoPs allows training from otel and your own internal sessions to understand SOE based on model
78
itsdad retweeted
The hypocrisy surrounding the Jensen Huang Scandal is remarkable. This is hit job. Large forces are working behind the scenes to get rid of Jensen just days before he was suppose to arrive in Shenzhen for APEC 2026. Jensen supports open source AI models and China leads the world in them! Go beyond the headlines and start digging to find the truth #JensenHuang
itsdad retweeted
Introducing the Gemini Decisions API
Powered by our new Gemini Decision 1 model, the API is built for fast routing, tool selection, safety checks, and grading inside your apps and agents without the latency or output-token cost of full text generation.
• Up to 1,000 questions evaluated per request
• $0.10/1M input tokens & $0 output tokens
• One shared input across text, images, audio, video, and PDFs
Start building today by grabbing an API key in @GoogleAIStudio
itsdad retweeted
Google open-sourced Harness².
(turn any harness into a self-improving system)
a harness is everything wrapped around the model, like the system prompt, tools, memory, skills, and subagents. today a person designs it once, and the agent runs every new task with that same fixed setup.
Harness² lets the agent refine that setup for each new task on its own, with the model weights left untouched.
on every task, the agent proposes edits to its harness, runs a few candidate versions, and composes a better harness by comparing how those runs behaved.
because it learns from the differences between runs, it needs no correct answer or grader for the task.
that covers a single task. Harness² also carries what it learns from one task into the next.
after each task, it saves the harness it settled on along with the lessons that held up across the runs. the next time it refines a harness, it starts from those notes instead of from scratch.
this means every finished task makes the next round of edits more informed.
Google tested this with Gemini 3.5 Flash and GLM-5 on legal, office, and professional work benchmarks. one round of edits was enough to beat their original harnesses, and giving the agent more rounds pushed scores higher.
those extra rounds can go wide or deep. the agent either tries several candidate harnesses at once or keeps refining a single one step by step, using OpenCode or Codex to carry out the work.
you can find the Harness² repo here: github.com/google-research/h…
itsdad retweeted
🚨 BREAKING: GEMINI SUBSCRIPTIONS ARE COMING TO HERMES. 🔥
Today, Gemini in Hermes runs on an API key, not the Gemini plan you already pay for.
That's being worked on, the approved way:
▸ Nous is working with Google to bring your Gemini subscription into Hermes
▸ It should be officially supported, with Google's sign-off
▸ Apps plugging in a Gemini plan today aren't doing it the way Google approves
▸ Not live yet, no date
Claude and ChatGPT subscribers can already bring their plan to Hermes. Gemini is next in line.
itsdad retweeted
Design with an elephant. Review with a goldfish.
Watch the 65-second @googledevs explainer on how the open source Elephant-Goldfish Model catches blind spots before AI agents write code.
👇 Video and blog post in first reply

ALT Dark-background title card for the 65-second Elephant-Goldfish Model (EGM) explainer video. At the top center is the minimalist split EGM mark: a blue elephant silhouette on the left facing an orange goldfish silhouette on the right. Centered white text below reads "Elephant-Goldfish Model" above the tagline "Design with an elephant. Review with a goldfish." ("elephant" in blue, "goldfish" in orange), with a bottom badge reading "Open source, from Google · egm.dev".
itsdad retweeted
How come no one told me about google's skill "Migrate Legacy Codebases to Safe Rust" github.com/google/skills/blo…
The whole security model of Agent Sandboxes fits in one argument.
The string is the code your agent wrote. external_lookup is the complete list of things it's allowed to call. No files, network or env vars unless you grant them.
uv run main.py 👇
pydantic.io/ih7Pn
itsdad retweeted
Tradovate is now integrated with OpenMarket.
Connect your account and trade futures directly from our charts, with footprints, heatmaps and order flow alongside every trade.
🤖 Made with AI
itsdad retweeted
Interesting take on the whole #JensenHuang trilogy
I wonder how much truth there is to that
itsdad retweeted
Everyone’s talking about Jensen Huang’s affairs. That’s not the real story.
Jensen Huang is out as Nvidia CEO. Colette Kress, the CFO he allegedly once called “a deer in the headlights,” is now running the company.
Here’s what else is coming out:
Nvidia’s commitments and guarantees jumped from $184 billion to $530 billion in just one quarter.
Some of these deals were allegedly pushed by Jensen’s own son-in-law.
A congressman’s wife allegedly sold her Nvidia stock right after it hit an all-time high, before it dropped.
Now the big question: is 20% of Nvidia’s revenue allegedly tied to chips illegally shipped to China?
The stock was falling before the public heard any of this. Coincidence? You decide.
Share this with someone who owns $NVDA.
#Nvidia #NVDA #Jensen
itsdad retweeted
🚨JUST IN: US intelligence can NOT prove Alice Huang spied for China, the case that cost Jensen Huang Nvidia $NVDA.
The intelligence gathered on her “does not prove” she was spying for Beijing, per CNN.
Elon Musk is now backing Huang’s return.
“A petition is underway to reinstate Jensen and I’m signing it too,” Musk posted.
Hours earlier he told Nvidia staff who want Jensen back to start with “NVIDIA is nothing without its people” and “resign and chill”.
Those are the words OpenAI staff used in 2023, when 505 of them signed a letter demanding their board resign and Sam Altman returned days later.
Inside Nvidia, an engineer had asked the board to weigh staff opinion first, and that post drew more than 1,600 comments.
Employees expected the investigation to finish before anything changed.
Huang resigned just as it began.
The FBI has been investigating for at least a year.
Officials at the Office of the Director of National Intelligence called the spy label “over-torqued”.
Attorney General Todd Blanche says there are “no conclusions reached by the Department of Justice”.
Trump was asked whether Huang was right to step down and said, “the big thing is, was the woman a spy? And I don’t know if that’s been determined yet.”
The first earnings without him land on November 17.
itsdad retweeted
终于实锤了❗️❗️
干掉黄仁勋的,不是婚外情,而是一场由美国政客、媒体、CIA和Anthropic 联合做局的“完美猎杀”。
黄仁勋辞职后,《纽约时报》两名记者在播客里亲口承认:
他们是收到了一位“科技界联系人”的爆料,才去查Megaspeed的。这个联系人,被开源情报指向了Anthropic 的联合创始人杰克·克拉克。
原来,这个Anthropic 的联合创始人杰克是彻头彻尾的恶魔😈。这个家伙后续必死无疑!
顺着资金链往下挖,全是对得上的:
Anthropic今年向支持“收紧对华芯片管制”的组织捐了 4000万美元。
逼黄仁勋辞职的参议员克莱默,收过Anthropic 5.1万美元政治捐款。
CIA监听黄乐一年拿不出间谍实锤,却在中期选举后精准放料。
纽时配合政客,用匿名员工爆料补上“第二段婚外情”。
CIA给黑料,Anthropic出钱,政客挥刀,媒体处刑。
黄仁勋反对封杀中国AI芯片,说“限制不明智”。
而Anthropic靠闭源API赚钱,最怕中国开源模型在GPU上免费跑。让英伟达戴上镣铐,Anthropic才能活下去。
他的“罪”,不是睡了谁。
黄仁勋挡了丑陋资本的路,挡住了美国政客的枪口,还是一个不听话的华人。
itsdad retweeted
Isn't Brian Mast the one who wears the Israeli military uniform to work, even though he says he's working for the United States? So, some foreigners are welcome to run spy ops against America and others are not?
itsdad retweeted
Your threats are ringing more and more hollow with the average person, and that terrifies you doesn’t it?
itsdad retweeted
Introducing: @browser_use x @1Password
Your agent can log in anywhere without ever seeing your password.
> connect 1Password once
> agent hits a login page
> you approve it in 1Password
> 1Password fills it in, agent keeps going
Try it 🔗⬇️

































