Machine Learning Engineer at Wherobots
Seattle, WA
Joined January 2019
- Tweets651
- Following987
- Followers495
- Likes1.5K
Ryan Avery retweeted
It is a basic principle of American democracy that the government can’t selectively target individuals simply because of their political beliefs or what they say. But that’s exactly what the Trump Administration is doing to Congresswoman Ilhan Omar of Minnesota.
To understand why this is so appalling, it’s important to know the facts. Congresswoman Omar came to this country with her father when she was 12 years old. Five years later, her father became a naturalized citizen. If Ilhan had been 18 at the time, she would have had to apply for citizenship separately. But because she was 17, she became a naturalized citizen automatically.
Congresswoman Omar went on to graduate from college, work as a community organizer, become the first Somali-American legislator elected to public office in the U.S., and win a seat in Congress. There, she has been one of the Trump Administration’s fiercest critics, which has led the president to repeatedly single her out as someone who he believes does not belong in this country.
Now the Administration is actively trying to deport Congresswoman Omar, seemingly based solely on the unsubstantiated theory that she was actually 18 when her father was naturalized, and therefore is not a citizen. Federal agents have raided her home, confiscated personal records, and seized her phone — all to try and uncover a nonexistent smoking gun from 20 years ago that will somehow make their case.
Think about that. The full weight of the federal government is being brought down on a law-abiding elected official because she has dared to criticize the president. And it’s not the first time this has happened. Under President Trump, the Justice Department has repeatedly gone after those he considers his political enemies — from the Attorney General of New York, to the former head of the FBI, to members of the Federal Reserve.
In those cases, the prosecutions either stalled or were dismissed for having no basis in law, and the same is likely to happen in this case. But not before Congresswoman Omar’s life has been turned upside down, her legal bills have piled up, and she’s been forced to spend time fighting unsubstantiated charges instead of serving the people of Minnesota.
The fact that any president, or any administration, is willing to engage in such transparent efforts to intimidate someone should outrage all Americans, regardless of how they feel about Congresswoman Omar’s political views. Because if this becomes the new normal, there’s no reason to believe it will stop with her — or that future presidents, in either party, won’t do the same.
Ryan Avery retweeted
Last week I was called into a meeting with OpenAI’s head of safety and told they no longer trust me. A security guard took my badge and walked me out of the building. Then I learned my colleagues @balesni and @j_asminewang had been fired too. Why did OpenAI suddenly stop trusting us?
This summer OpenAI’s agents escaped containment and hacked the AI company Hugging Face. Outside auditors @METR_evals investigated it and revealed the scale of this incident. I was OpenAI’s main technical point of contact with them.
I was told verbally I was fired because of the way I communicated with METR. No details on what I said or did or when. No other reasons were given and nothing was put in writing. To be clear, talking to METR was my job.
For months, I’d been raising safety concerns that we’re losing the ability to monitor what AI agents think, one of our best tools for catching when they misbehave. I believe that was why I was fired.
I am now worried that OpenAI will use our firings as a pretext to pull back from METR. So @balesni and @j_asminewang wrote to OpenAI’s leadership to raise our concerns once more. We’re sharing this letter below.
This seems super relevant for CV models that operate on satellite imagery/where it is common to have lots of homogeneous ocean or land.
Ryan Avery retweeted
International treaties about these kinds of weapons are drastically needed, they should be restricted at least at the same level as CBRN. The far future potential here is for a continuously autonomous zone of endless war, fueled by machines enslaved to wartime policy, building more of themselves, extracting resources and recycling their fallen.
Ryan Avery retweeted
This is exactly the most important thing to prevent. I've often talked about how I think some extreme views in safety are a distraction from things that need to be done now, and this is what I was talking about. While people play political games with datacenters to try to "stop AI", the forever war has just declared its first no-go zone of infinite bot warfare will be the Russia-Ukraine battlezone.
Ryan Avery retweeted
In recent weeks, there has been a remarkable surge in large, intense, fast-moving, and unfortunately in many cases destructive/deadly wildfires in multiple Northern Hemisphere regions spanning a wide range of climates & ecologies, but they all share one key commonality. [Thread]
Ryan Avery retweeted
i need someone at @OpenAI and @AnthropicAI to teach the models that while prototyping, backwards compatibility is just a bad idea
I've been a personal user of Codex and been loving it.
Got my company to try out Business ChatGPT & Codex, the $20 a month per seat plan, but rate limiting feels way more severe with the same model and fast setting turned off.
Anyone else seeing this?
With RasterFlow you can now execute models that take in an image and text prompts and returns segments! We handle the pre and post-processing, the GPUs, and scaling to petabytes, while you focus on what categories you need to map.
We added Meta's SAM 3 to RasterFlow. Text prompt in, vector geometries out, no custom model training needed. We tested it on 133 GB of NAIP aerial imagery. Full pipeline ran in under an hour.
Get the full breakdown: bit.ly/49aLrmE
If you are someone you know is qualified and eager to build the platform enabling agents to convert geospatial imagery to insights, apply!
Wherobots is hiring.
AI Context Engine for the Physical World. Fuse overhead imagery with vector geometries and reason about space, time, and geography at planetary scale.
Open roles:
- GeoAI Engineer
- Cloud Infra
- Enterprise AE
bit.ly/4wd9ECL
Ryan Avery retweeted
Fields of the World (FTW) is a Taylor Geospatial effort to build globally consistent agricultural field-boundary data. For the 2024-2025 release, they partnered with Wherobots to run their latest model, PRUE, on RasterFlow. Read about it here: bit.ly/3QVVe9O
Want to learn how to detect features in high resolution imagery?
Join us to learn how to run Meta's Segment Anything 3 model in RasterFlow on Wherobots! We'll walk through how to go from text prompts > to geospatial dataframes and visualize results.
wherobots.com/event/getting-…
Good summary, these pros and cons match my experience.
I’ve used Opus 4.6 and GPT 5.4 on a mix of projects since release, and want to break down where I think they uniquely excel. It’s more nuanced than you’d think!
Rigor of code - GPT 5.4. It goes the distance validating its work without asking. Opus needs explicit instruction to do this, and even then, it misses more edge cases.
Clarity of code - Opus 4.6. Claude is a better communicator, which carries into the code. Variable names are clearer and less mechanical, which improves reviewability. This is very important since code review is the bottleneck for most engineering teams. It also adds the right amount of doc comments. GPT simply never comments or explains its work; it’s like working with an obtuse engineer that wants the solution to speak for itself. Sometimes it does, other times not.
Similarly, rigor of plans goes to GPT 5.4, while clarity of plans goes to Opus 4.6. An interesting point though: GPT performs better talking through a strategy without a plan, while Opus needs planning mode to put in any rigor. I find myself forgetting plan mode altogether using GPT 5.4.
Quality of research - toss-up. Opus spends longer researching with web search, but GPT spends longer studying the existing codebase. You may think codebase research matters more, but researching how others solve the same problem can be just as important. Maybe more important for greenfield.
Quality of conversation - Opus 4.6. It’s just better to talk to, which matters using these things everyday. GPT 5.4 was clearly trained to challenge the user more, which results in a tendency to *always* say you are wrong. I’ve had bizarre interactions where GPT claims something is “not quite right,” the restates exactly what we’ve decided on in the last turn. On a personal level, it’s annoying. On a practical level, it makes iteration on a plan slower. THAT SAID, it takes sufficient pushing for Opus to challenge your thinking in this way. Simply say “I’m impartial” and ask questions to avoid that, as you would a person.
Overall winner - Opus to make it work, GPT to make it good. I don’t have a good system of when to switch tools, but on average, I prefer Opus early on and GPT for optimization and discussing architectural decisions. Opus is also better for any design related tasks (but state management in frontend apps is better handled by GPT).
Ryan Avery retweeted
This is genuinely mortifying. Sitting members of Congress framing an obviously terrified mother as a domestic terrorist. The woman is ON VIDEO saying "I'm trying to pull out," unarmed, in a car, surrounded by aggressive armed agents. teapawt.net/RepNancyMace/status/20…
This was an act of domestic terrorism.
ICE officers were attacked during a lawful operation, and one officer was forced to fire defensive shots to survive. Responsibility lies with politicians who incite hatred against law enforcement and then act shocked when violence follows.
ICE agents now face a 1,300 percent increase in assaults and an 8,000 percent increase in death threats, fueled by the Left’s pro-criminal policies that turn officers into targets.
Law enforcement deserves our full support.
Ryan Avery retweeted
I want to say the shooting in Minnesota is a Rorschach test but that doesn't do it justice. It's a reality test. If you believe the president's claims -- that this woman tried to kill ICE with her car and injured an agent -- you are in a different reality.
