The future is bright.
California
Joined June 2007
- Tweets20K
- Following724
- Followers152K
- Likes31.1K
Matthew Berman retweeted
🚨 Sam Altman Says Anthropic’s Push to ‘Ascribe Personhood’ to Their Models is a ‘Dangerous Path’
“I hope we come to an understanding that this technology is not a person and is certainly not superior to people … Even the people building AI get confused about that.”
Matthew Berman retweeted
OpenAI CEO Sam Altman:
"the biggest advantage we have is that we get the next generation of models before anybody else, and we have time internally to use them and see what's going to work"
Take the math results. I don’t think any model currently widely available can get close to that
My interview with Sam Altman (@sama):
0:00 Intro
0:52 When Will AI Discover New Treatments?
3:31 Why AI Hasn’t Killed Coding Jobs
11:47 Personal AI Assistants
15:39 Why Are We Busier Than Ever?
20:10 Why the AI Consciousness Debate Worries Sam
26:09 ChatGPT Is Too Agreeable
30:03 OpenAI’s Bet on Chips and Cheaper Compute
34:20 Why AI Needs a New Kind of Computer
37:41 Sam on Pressure: “Pick the Big Thing”
42:18 Why “AI Will Cure Cancer” Isn’t Enough
Matthew Berman retweeted
“I don't think we have full information.”
Christian Catalini on why the AI safety debate needs better evidence. From our September 2026 interview.
Matthew Berman retweeted
OpenAI's Tibo reveals he has ADHD and thrives on context-switching but says Ultrafast is delightful when he wants to focus on a single thing
Tibo: People have different kinds of preferences on whether they like to be, you know, mono-threaded, as we talked about, or multitask a lot. For folks who like to multitask a lot, you don't benefit as much from Ultrafast.
Berman: Some people don't like to change context all the time. Where do you fall on that spectrum?
Tibo: I have ADHD, so I context-switch all the time.
Berman: It's funny, 'cause I also have ADHD and I actually don't want to context-switch all the time. That's really hard for me. I want to focus on two to three, and that's why I was so excited about Ultrafast. That's fascinating that you're the opposite there.
Tibo: You know, I thrive in context-switching and making lots of little decisions. But sometimes I do want to just stay focused on one thing, and then Ultrafast is just delightful, because it just keeps you right there in the flow.
Dylan Patel reveals that a friend at Anthropic stopped taking her ADHD medication once Mythos was good and available internally, because being ADHD made her better at managing AI agents
Dylan: "I was asking because I have a friend at Anthropic, and the moment Mythos was good and available internally, she told me that she stopped taking her ADHD medicine."
Jordan: "Oh, come on."
Dylan: "And that made her a better employee."
Jordan: "It made her a better employee."
Dylan: "Yes. Because she was able to manage the agents and context switch and be ADHD."
Jordan: "How is she as a friend?"
Dylan: "Oh, she's a great friend."
Matthew Berman retweeted
One of the most interesting things Tibo said:
laptops were designed around human constraints. AI models don’t have those same constraints.
So why should future agents be limited to the computer sitting on your desk?
We’re already seeing the beginning of this. ChatGPT Work can operate inside its own cloud computer instead of depending on your hardware.
Now extrapolate that a little:
→ persistent cloud machines
→ huge amounts of RAM/storage
→ many agents working in parallel
→ specialized tools and environments
→ tasks running for hours or days
→ compute scaling up whenever the agent needs it
The “AI computer” of the future may not look anything like a laptop.
Your laptop might just become the window you use to talk to it.
Matthew Berman retweeted
LMAO the Codex reset lore somehow gets even better
Tibo says the resets aren’t coordinated with marketing or finance
He can literally reset everyone’s usage whenever he thinks it feels right
AND THERE IS NOW AN ACTUAL PHYSICAL RESET BUTTON 😭
OpenAI may have accidentally invented the most powerful button in AI
My full interview with Tibo (@thsottiaux)
0:45 Tibo's Lessons from Google DeepMind
4:22 Building OpenAI’s Relentless Culture
7:23 Astra & Next Gen Models
11:18 How Fast AI Changes Developer Workflows
14:27 ChatGPT & Codex Merging
20:25 OpenAI vs. Anthropic
23:37 Why OpenAI Keeps Resetting Limits
30:25 Recursive Self-Improvement
32:00 Dangers That Caused "The Pause"
34:13 Will Ultra Fast Become the Default?
43:20 Why Everyone Needs to Try AI
Matthew Berman retweeted
here are my favourite parts from the tibo interview 👀
> outside ultrafast, the speed you get is roughly 60% faster than 3 months ago, and they're going to publish how
> sol is significantly more token efficient than terra, and the next model (assuming astra) will be significantly more token efficient than sol
> he thinks ultrafast speeds could be the default, or close to it, in a year or two
> openai employees don't all get ultrafast, the vast majority of that capacity is reserved for customers
> he admits the current harness is a bit clunky: skill files are hard to maintain, memory doesn't always remember, subagents can break the illusion
> the reset button isn't coordinated with marketing or finance, and yes, there's a physical button, "i can press the button whenever i want, whenever it feels right"
> chatgpt and codex merged because the next models are one agent that does everything, "you and your mom will use the same thing"
My full interview with Tibo (@thsottiaux)
0:45 Tibo's Lessons from Google DeepMind
4:22 Building OpenAI’s Relentless Culture
7:23 Astra & Next Gen Models
11:18 How Fast AI Changes Developer Workflows
14:27 ChatGPT & Codex Merging
20:25 OpenAI vs. Anthropic
23:37 Why OpenAI Keeps Resetting Limits
30:25 Recursive Self-Improvement
32:00 Dangers That Caused "The Pause"
34:13 Will Ultra Fast Become the Default?
43:20 Why Everyone Needs to Try AI
Matthew Berman retweeted
JUST IN: Demis Hassabis is stepping down as CEO of Google Deepmind AND Jeff Dean (legendary computer scientist) is also leaving to start his own company. - Axios
Matthew Berman retweeted
One of the most fun conversations I have had in a long time. Thanks for having me on your pod @MatthewBerman. I could’ve gone for a few hours. So fun man!
> "In no scenario is it bad for the US to have frontier-class open-source models."
> "Every time there’s a step-function improvement in AI, a human bottleneck emerges alongside it."
> "Your desire to be liked sometimes takes away from your desire to do the right thing and seek the truth."
> "Every job will get refactored. The people who aren’t AI fluent will have a very hard time finding jobs."
Jeetu Patel (@jpatel41), President & Chief Product Officer at Cisco sat down for an expansive talk about life and the future.
00:00 AI, jobs, and human potential
00:41 AI cybersecurity and Cisco’s security models
06:33 Why Cisco released the models openly
08:08 Running AI models locally
10:12 Why people are negative about AI
12:14 Will humans always be needed?
14:56 Can curiosity and ambition be taught?
17:12 Passion vs. expertise
19:39 The importance of unlearning
23:52 How to find the truth
26:06 How to get people excited about AI
33:35 How open source affects the AI industry
35:07 Can open-source AI make money?
46:44 Are Chinese open-source models good for the US?
49:55 Should the US restrict Chinese AI models?
53:05 Does open source make humanity safer?
54:11 Advice for people worried about AI
Matthew Berman retweeted
We are announcing a long-term strategic partnership with NVIDIA. NVIDIA is making a substantial investment in SSI that will let us 10x our compute in the next 12 months. We reached the point where our research is worth scaling and with this partnership we will be able to. We are honored by NVIDIA’s conviction.
Matthew Berman retweeted
Google CEO Sundar Pichai discusses how AI agents might impact the Internet in the future.
"It will allow humans to use the Internet in ways that give them joy & purpose, & not always be forced to fill out [a bunch of form fields to renew a DMV license, for example]."
Video v/@googledevs & @MatthewBerman
Matthew Berman retweeted
Meet LM Studio Bionic.
The Agent made for Open Models.
lmstudio.ai/bionic
Matthew Berman retweeted
Google's Chief Business Strategist for AI @GopiKallayil argues AI outranks every tool humans have built across 12,000 years of problem-solving.
"There has never been a tool that can give one person the thinking power of many...until now."
Link below.
Matthew Berman retweeted
My strangest hobby has always been thinking about other people’s businesses.
AI has made that habit way more useful. Every company has workflows, bottlenecks, and weird internal habits that are worth looking at again.
I’m doing a small number of free sessions in partnership with @ForwardFuture.
Request a session here: forge.forwardfuture.ai/
.@hnshah and @ForwardFuture are partnering to offer free consulting sessions with select enterprise companies.
Space is very limited.
If you are looking for help implementing AI, apply here:
forge.forwardfuture.ai/
Matthew Berman retweeted
My thoughts on the future of model routing and AI:
- We have not even scratched the surface of runaway inference costs
- Solving this requires intelligent model routing, especially as the inference landscape continues fragmenting. This is a *hard* problem.
- Naive solutions (turn-based routing, session routing) fail; routing successfully involves managing multiple cost surface areas in concert.
- Getting routing right means a more diverse market of providers, more power for consumers, reduced ecological impact, and improved effectiveness.
More in the full essay:
I made an angel investment in @tomas_hk two years ago, believing model routing would become a massive business.
Now companies are spending $500m per month on tokens. Increased quality and decreased cost are not mutually exclusive.
Here's Tomas' essay about how Model Routing Will Control the Future of Economic Value
briefing.forwardfuture.ai/p/…
Matthew Berman retweeted
When a system mediates what you read, what you write, what you remember, and what options you even notice, understanding that system stops being a hobby and becomes literacy
From my article on Matthew's AI Newsletter
Ahmad (@TheAhmadOsman) wrote an incredible essay about the importance of having your own compute architecture.
Local LLMs, Buy a GPU, and the Case for Cognitive Security
briefing.forwardfuture.ai/p/…
