Red (@adventurared) header
Red (@adventurared) profile picture

🇲🇽🏳️‍🌈 bytes.🧑🏽‍🔬 connecting people to develop artificial intelligence @openai. my tweets don’t represent anyone/thing else

San Francisco, CA
Joined July 2012
Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)Photo by Red (@adventurared)
I see five fundamental issues when it comes to street conditions in San Francisco. These negative street conditions are largely driven by a relatively small group of regular public drug users (1,000–4,000 people out of 17,000–30,000 fentanyl users in the city). These public drug users should be strictly distinguished from the homeless, the vast majority of whom are neither visible on the street nor active drug users. First, drug treatment has extremely low success rates and, even with tons of investment, is not significantly decreasing the amount of drugs being consumed in the city. Treating drug addiction is hard. Treating drug addiction in a place with cheap and readily available drugs is harder. Second, there is a public policy disagreement about the role of the police in the fight against public drug use. As a result, we have almost no tools to disincentivize public drug use. We have effectively decriminalized drug use in California by making it a misdemeanor. The vast majority of the time, the only thing police can do is tell users to move along or give them a ticket. The RESET Center allows police to arrest users and remove them from the street, but it’s only available in a tiny part of the city. Outreach workers have no power to compel public drug users to stop using. Third, we have widely available drugs that are cheaper than in almost any other city in America. The cheap drugs draw users from cities and towns across the region and the country. Drug users don’t come to San Francisco for our services. They come here for our cheap drugs. Fourth, there are no good tools for reducing supply or increasing cost in our city. Fentanyl is manufactured at scale from industrial chemicals that are cheap and easily available. It’s so potent that San Francisco is estimated to consume only 200–400 pounds of pure fentanyl a year. That is too easy to smuggle in. The labor supply of dealers is effectively infinite: they make $3–4 a day where they come from and $400–700 a day here. Arresting dealers doesn’t reduce the availability of drugs or increase the price. We’ve had more than 3,000 dealer arrests in the past three years. It’s changed how dealers deal, but drug consumption is actually up, not down. Fifth, we have a Housing First policy that provides temporary and permanent housing to thousands of active drug users in very concentrated parts of the city. As a result, we create the high-drug-use areas that dealers target. Residents of those neighborhoods are held hostage by the drug trade and rampant public drug use. Your tax dollars create high drug use neighborhoods. San Francisco city government understands all of these facts. All of this information is public and findable. Nothing will be done unless you understand these facts and demand real solutions. Every time you ask for ineffective solutions based on a lack of knowledge, you give your elected officials license to spend your money on programs/policies that will never work.
89
58
14
837
110,534
I bet you don't know what a SF "public bank" would be. absolutely no offense, very few do. seriously - ask your friends. "um, it's a place for residents to get fee-free checking accounts, low-rate mortgages, etc?" nope. not even close. and Prop B proponents don't seem to want you to know the details... well here they are. hot off @theVOSF press. enjoy! @JackieFielder_ @bilalmahmood @alankennywong @scsherrill @MHurabiell @SFNoPropB @sfpublicbank thevoicesf.org/public-bank-d…
21
24
2
137
6,817
Red retweeted
we made a chip and it is fast
3,937
1,988
1,419
47,563
8,610,669
I asked GPT-5.6 Sol to create the most Claude-y possible parody image and what it came up with is pretty great and dead-on.
🤖 Made with AI
83
78
23
1,554
307,408
We integrated the Codex harness into Prism (prism.openai.com) — this means you get skills, reasoning levels, and the raw tenacity of the Codex model in your LaTeX environment. Oh and we also built version mgmt into Prism, which was one of the top requests. See below thread from @vicapow for some great examples of the power of Codex inside Prism 👇
🧵1/ We've brought the most advanced AI to Prism by introducing Codex to Prism. Prism is already the best place for scientific writing to happen—and with Codex, now you can write, compute, analyze, and iterate all in one place.
8
21
5
319
42,563
>be you >work in HFT shaving nanoseconds off latency or extracting bps from models >have existential dread >see this tweet, wonder if your skills could be better used making AGI >apply to attend this party, meet the openai team >build AGI
540
308
242
9,197
2,065,827
>be me >grind for a decade trying to help make superintelligence to cure cancer or whatever >mostly no one cares for first 7.5 years, then for 2.5 years everyone hates you for everything >wake up one day to hundreds of messages: "look i made you into a twink ghibli style haha"
2,253
2,241
1,260
50,737
5,057,829
if you are a safety researcher, please consider applying to help test o3-mini and o3. excited to get these out for general availability soon. extremely proud of all of openai for the work and ingenuity that went into creating these models; they are great.
487
515
117
8,234
806,290
OpenAI comms: [underspecific hype-y 'big tings coming!!' pls like and subscribe] Google comms: [corporate vagueness about Gemini3-0011 v2 FINAL.docx on Vertex available to 14 users] GDM comms: [we have simulated a rat's brain capable of solving 4D chess, but we're not sure why] Anthropic comms: [we are very worried. we have asked the model to spell out 'doom' and it did] Meta comms: [haha we love the USA!! our new mostly-open model is very American!! pls work for us] Microsoft comms: [would you like to set Microsoft Edge your default browser?]
39
160
28
2,241
248,044
🌐 Introducing ChatGPT search 🌐 ChatGPT can now search the web in a much better way than before so you get fast, timely answers with links to relevant web sources. openai.com/index/introducing…
901
2,556
1,683
15,316
4,526,702
everyone wants to join a rocket ship until they remember they’re bumpy and on fire. buckle up y’all, only way to space.
23
71
14
890
159,026
First @NVIDIA DGX H200 in the world, hand-delivered to OpenAI and dedicated by Jensen "to advance AI, computing, and humanity":
808
2,037
970
21,369
4,402,695
I got a rare opportunity to ask OpenAI's video generation model Sora for some videos this week! I had 3 prompts. Here's the videos with my prompts and what I learned 🧵
374
1,813
204
29,010
7,369,357
scaling laws are decided by god; the constants are determined by members of the technical staff
865
643
178
9,886
1,613,090
My typical day as a Member of Technical Staff at OpenAI: [9:00am] Wake up [9:30am] Commute to Mission SF via Waymo. Grab avocado toast from Tartine [9:45 am] Recite OpenAI charter. Pray to optimization Gods. Learn the Bitter Lesson [10:00am] Meetings (Google Meet). Discuss how to train larger models on more data [11:00am] Write code to train larger models on more data. pair=@hwchung27 [12:00pm] Lunch at the canteen (vegan, gluten-free) [1:00pm] Actually train large models models on more data [2:00pm] Debug infra issues (why the fck did I pull from master?) [3:00pm] Babysit model training. Play with Sora [4:00pm] Prompt engineer aforementioned large models trained on more data [4:30pm] Short break, sit on avocado chair. Wonder how good Gemini Ultra actually is [5:00pm] Brainstorm potential algorithmic improvements for models [5:05pm] Conclude that algorithmic changes are too risky. Safer to just scale compute and data [6:00pm] Dinner. Clam chowder with Roon [7:00pm] Commute back home [8:00pm] Have a wine and get back to coding. Ballmer’s peak is coming [9:00pm] Analyze experimental runs. I have a love/hate relationship with wandb [10:00pm] Launch experiments to run overnight and get results by tomorrow morning [1:00am] Experiments actually get launched [1:15am] Bedtime. Satya and Jensen watch from above. Compression is all you need. Good night
137
293
100
3,635
908,855
openai is the most talented and nicest group of people i have ever seen in one place working on the hardest, most interesting, and most important problems with all the key resources in place extremely focused on making AGI you should perhaps considering joining us
1,696
939
316
21,095
3,668,074
Replying to @sama
“A half duck half dragon flies through a beautiful sunset with a hamster dressed in adventure gear on its back”
188
256
118
4,019
1,374,991