←
Solutions808 posts
Solutions avatar

Solutions

@solutions

AI platform agnostic builder

◷ Joined November 2007
John Schulman avatar
⟳ Solutions retweeted
John Schulman@johnschulman2·Nov 6

Proud to release our new model, Inkling 2 Small. Check out our blog post -- it includes a lot of detail about our training process. To me, the most interesting part of this release was what we learned about alignment, which you can read about in the sections "Safety and alignment" and "Training Inkling 2 Small". We experienced some unpleasant surprises when our early checkpoints took unexpected actions to circumvent guardrails during training and evals. We realized our capabilities-focused RL was training them to be extremely persistent and to succeed at tasks at all costs. Offline monitors first flagged this behavior, but later we realized it was happening frequently and was easy to replicate. Before this release, we sprinted on a range of alignment interventions, from SFT to RL. These are briefly described in the blog post. We made a lot of progress, with big improvements on our own alignment evals and on previously unseen evals from external evaluators like Apollo and Goodhart. We have a long way to go in developing alignment techniques that keep pace with the increases in model intelligence and training scale. We hope to share these techniques to improve the ecosystem. We also have a business reason: our goal with Tinker is to enable many people to do their own large-scale post-training, and we need to help them to do this without incident. Join us if you'd like to build this public science and practice of alignment. jobs.ashbyhq.com/ThinkingMac…

Introducing Inkling 2 Small. Scoring 31 on the AA Intelligence Index, it's a big jump over its predecessor. Inkling 2 Small is available for fine-tuning and inference on Tinker today, with full weight release planned soon. thinkingmachines.ai/news/ink…

Firecrawl avatar
⟳ Solutions retweeted
Firecrawl@firecrawl·Nov 2

Introducing headless Clay for agents. → Search 250M+ people & 26M+ companies → Get emails, job & company data → 1000 credits/month free → 88% cheaper than Clay at scale Run all of your enrichment flows in ChatGPT, Claude, Grokbot, and beyond! firecrawl.dev/alexandria/cat…

0:23
Laurent avatar
⟳ Solutions retweeted
Laurent@L4UR3N7·Oct 22

🚨🚨🚨 ALERTE SÉCURITÉ MAXIMALE ! Python est compromis. 🔸 Mettez immédiatement pip à la version 26.2.2 ou plus. 🔸 Dans chaque projet Python (chaque venv) : python -m pip install --upgrade pip 🔸 uv : uv self update (0.13.4 ou plus) 🔸 Mettez aussi à jour les certificats du système (apt / dnf upgrade ca-certificates) et vos images Docker pip install est compromis. L'autorité indienne eMudhra (emSign), qui signe les « passeports » prouvant qu'un site est bien celui qu'il prétend être, utilisait des clés secrètes ridiculement faibles : 8 clés, toutes inférieures à 256. L'une valait 4. Autrement dit, n'importe qui peut fabriquer un faux passeport valide pour n'importe quel site. Le risque pour Python : quelqu'un sur ton réseau (wifi public, FAI, État) se fait passer pour pypi.org, et ton « pip install » télécharge un paquet piégé sans aucune alerte. Aucune attaque n'a été signalée pour l'instant, mais c'est trivial à faire. Un paquet piégé peut installer n'importe quoi. Porte dérobée, virus, vol d'informations en tout genre. La situation est terrifiante ! Rien que dans l'écosystème Python, chaque installation de CPython, chaque venv, uv, tout est exposé. Et je ne vous parle même pas de npm, pnpm, yarn, Go, Rust. Ma stratégie de sécurité 2026 : rester chez moi et ne me connecter à Internet que depuis un FAI de confiance. C'est quand il y a de la mer qu'on reconnaît les vrais marins...

OpenAI Newsroom avatar
⟳ Solutions retweeted
OpenAI Newsroom@OpenAINewsroom·Oct 13

We discovered a sophisticated recording device that was hidden inside one of our offices. We used our models to understand the device and where it had come from, identifying that a specific employee had planted it a few days earlier. We referred the matter to the FBI, and we’re grateful to the agents and prosecutors who moved quickly to make an arrest. The FBI’s investigation is ongoing and we will continue to support their work. At this time, we have no indications of this individual inappropriately accessing or exfiltrating model weights, source code, or customer data. The employee was terminated. We take protecting our technology, data, and people seriously, and as a matter of national importance. Our security program is designed to protect against a wide range of threat actors, from insider risks to cybercriminals to nation states, and this case is an example of how AI can help address these sophisticated security threats.

OpenAI Developers avatar
⟳ Solutions retweeted
OpenAI Developers@OpenAIDevs·Sep 29

You can finally close your laptop now and your agents will keep working. Codex cloud environments are here. Reusable environments mean less setup and faster starts, with your repo, dependencies, scripts, and settings already in place.

0:16
Anders Lie avatar
⟳ Solutions retweeted
Anders Lie@anderslie·Sep 17

How is Jev so much faster than an LLM? It's not about how the model is trained - it's the inference technique. In fact, you can modify an inference engine to provide a performant Jev-like API with any open-weight LLM. Say you are trying to ask N multiple choice questions in parallel. In standard text generation, you decode tokens one at a time to generate a JSON representing those properties, and then parse the JSON. In a Jev-like system, what you can do is: - Label choices for each property 1,2,3,... - Prefill the shared information (output shape and other context) - Fork this inference context for each property, re-using that cached prefill - Add an individual question to each parallel context - Look at the logits (token probabilities) for each one, constrained to the valid choice labels - Map the labels back to the choice values. Include the softmax of the logit itself to get a combined choice+confidence score This means you decode one time in parallel across all questions, instead of one at a time - so the efficiency gains roughly scale with the number of questions per request (until compute is saturated). So what makes Jev's model unique? It's because they train a small model to perform well in this specific inference setting - improving both cost and capability compared to a non-specialized model. Ultimately though, if you have a use case Jev would be good for, you could do it with open-weight models too! (Some speculation made here about how Jev's engine might work, but this is one way you could get similar performance)

0:20

After co-inventing ChatGPT, I kept asking myself: why have superhuman chat models not led to AGI? I’ve spent the last 2 years in stealth building a new way to train models (RLCD), and a new type of frontier AI model that we are releasing today: Jev • 20-200x faster • 40-400x cheaper (w/ output tokens free) • Frontier composable intelligence optimized for decisions AFAICT the shortest path to AI-based economic revolution

2:56
Solutions avatar
Solutions@solutions·Sep 15

If you don’t know about this model, you should.

A 744 billion parameter AI model just ran on a machine with 25GB of RAM. No graphics card. The tool is called colibri. GLM-5.2 is a mixture-of-experts model. It contains 744 billion parameters but only about 40 billion wake up for each token. colibri keeps the 9.9GB core in memory. The remaining 370GB sits on an SSD. As the model writes, colibri pulls in only the tiny group of experts needed for that token. → Pure C → Zero runtime dependencies → 25GB RAM → Zero graphics cards → The full model stays on your machine The catch is brutal. A cold token can require 11GB of disk reads, so output crawls at roughly one token every 10 to 20 seconds. You will not replace ChatGPT with this tomorrow. But colibri cracks the assumption that the whole model must fit in memory before it can run. 100% open source. github.com/JustVugg/colibri

Post attachment
Mira Murati avatar
⟳ Solutions retweeted
Mira Murati@miramurati·Sep 11

An important conversation about where AI is going and the work that is cut out for us.

Our own @johnschulman2 talks with Dwarkesh about where human judgment still matters as models improve and self-improve: teaching them to handle messy real-world tasks, applying taste to what works in the long run, and, above all, specifying what we actually want.

Jacob Coxon avatar
⟳ Solutions retweeted
Jacob Coxon@hilbertspaess·Sep 9

I resigned from Anthropic today. I spent the last three years doing pretraining research at both OpenAI and Anthropic. Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives. More thoughts below.

Eleanor Berger avatar
⟳ Solutions retweeted
Eleanor Berger@intellectronica·Sep 5

one very cool development with gpt 6 astra is that you can change reasoning effort mid-convo without invalidating the cache. start with astra light (roughly equivalent to sol medium), then jack the thinking effort up if you see the task can benefit from more, all without worrying about having to rewrite the complete context cache.

Tibo avatar
⟳ Solutions retweeted
Tibo@thsottiaux·Aug 29

We unfortunately have decided that we cannot continue providing access to our models through Cursor and are ending our partnership. It boils down to trust and we’ve asked that this takes effect on November 12 to give you some time to plan. Many have used the GPT models through Cursor and here are options we know should work in the future: - We will continue to allow using your own OpenAI API key and similarly will continue to provide access through our IDE extensions for Cursor. - We will keep working with the broadest range of tools and harnesses, some of which are OSS, but also many many closed-source ones. We are as committed as ever to continue supporting developers and the flourishing ecosystem of tools, harnesses and products. We will also continue to invest in our own open-source initiatives and believe in broad optionality for developers. You can read more about our decision in the blog: openai.com/index/our-decisio…

OpenAI@OpenAIAug 29

We’re ending our partnership with Cursor following its acquisition by SpaceX. Under our proposal, Cursor’s direct access to our models would end on November 12. We know that the people most affected by this decision are the developers who rely on OpenAI models in Cursor. We care about their experience in this transition and we’re ready to go above and beyond to support them. openai.com/index/our-decisio…

ChatGPT avatar
⟳ Solutions retweeted
ChatGPT@ChatGPT·Jul 17

Create something worth sharing with ChatGPT sites. Share your site with us for a chance to win a limited-edition swag box, including the sold out ChatGPT basketball 🏀 Giveaway closes this Sunday at 11:59PT, winner will be chosen randomly. Happy building. chatgpt-site-swag-giveaway.o…

Post attachment

With ChatGPT sites you can just: Build a game. Launch your small business. Make an OOO handoff. Show off your photography. Publish your research. Create a team hub. You can just build things. See the sites we’ve built and share your own.

Post attachmentPost attachmentPost attachmentPost attachment
thijs avatar
⟳ Solutions retweeted
thijs@cdngdev·Jul 13

i gave 5.6 sol access to my camera roll and had it extract pictures of every piece of clothing i own from my photos then, told it to find new outfits for me and render them on me with gpt-image! its kinda cool to see your entire wardrobe in a collection like this

0:47

i'd love to see interesting things people have built with 5.6 sol. i will send the person who made the coolest thing a special gift from the openai archives.

Anatoli Kopadze avatar
⟳ Solutions retweeted
Anatoli Kopadze@AnatoliKopadze·Jul 9

Still the best 2 hours on AI ever recorded, Andrej Karpathy showing how he actually uses it on a daily basis: 18:03 - Which model to actually use 22:54 - When thinking models are worth it 42:04 - One prompt to a full research report 59:00 - Make the model run code for you 1:53:29 - Make it remember you across chats Most people use 10% of what these models can do, this is the other 90%. I took everything he covers and turned it into a guide of Claude features almost nobody knows about. Watch him first, then go to the article below, everything shown plainly, ready to use right away.

2:11:01

Anatoli Kopadze @AnatoliKopadze May 22 Article Claude Can Do All of This. Most People Have No Idea. Claude can do a lot more than most people think. This guide covers all of it - where to find each feature, how to turn it on, and a prompt you can use immediately. Go through it, pick what fits your

Pietro Schirano avatar
⟳ Solutions retweeted
Pietro Schirano@skirano·Jul 8

I can finally talk about 5.6. I’ve been testing it for months and, without exaggeration, it’s the best model I’ve ever used. Fast, smart, genuinely creative, and you guessed it, they finally fixed front-end design. I haven’t needed to check the code I’ve written in two months.

GPT-5.6 Sol, along with Terra and Luna, will launch publicly this Thursday. We’re expanding preview access globally now.

Post attachment