Diego Aud @dieaud91

Biologist & Nutritionist 🧬 | AI Enthusiast 🤖 | Exploring the intersection of biology and AI. Sharing insights on #AI, #Biotech, #Futureofwork Italia Joined November 2023

Tweets

4K
Followers

709
Following

603
Likes

45K

Diego Aud @dieaud91

15 minutes ago

I'm increasingly appreciating Anthropic's approach: less hype, more focus on shipping frontier models as fast as possible. My guess is that OpenAI's huge free user base is pushing it to prioritize compute efficiency. Leaked Terra 6.1 feels like a reflection of that.

Diego Aud @dieaud91

20 minutes ago

@rezoundous Meh. I’d have happily taken 28 days of radio silence if it meant getting Astra 6.1 by the end of October.

0 0 1 48 0

0 0 3 38 0

View Details

Diego Aud @dieaud91

20 minutes ago

@rezoundous Meh. I’d have happily taken 28 days of radio silence if it meant getting Astra 6.1 by the end of October.

0 0 1 48 0

View Details

Diego Aud @dieaud91

27 minutes ago

@artuskg 💀💀 And starting November 1st, my usage allowance on the $200/month plan is effectively getting cut in half...

1 0 2 7 0

View Details

Diego Aud @dieaud91

17 hours ago

What is OpenAI even doing? A Terra-class model (so even smaller than Sol) as an answer to a next-generation Fable/Mythos-class model?! As an OpenAI customer, I don't know what to think anymore. I'm ready to switch subscriptions if Fable 5.5 clearly outperforms GPT-6 Astra. Sorry, but I only use frontier models for my work.

Angel 🌼 @Angaisb_

21 hours ago

Not a new Terra model to compete with Opus 5.5 🥀 I hope it means we're also getting GPT-6.1 Astra because it'd be so disappointing if not

15 0 227 9K 12

4 2 16 1K 0

View Details

Diego Aud @dieaud91

35 minutes ago

@artuskg Yeah, that's why Fable 5.5 (hopefully dropping this week) might tempt me to get a Claude sub. Unfortunately, I'm not sure I can afford another subscription costing hundreds a month... Right now, my only other sub is Google Pro, and Gemini 4 Argon feels pretty good.

1 0 2 77 0

View Details

Diego Aud @dieaud91

6 hours ago

I get the idea. If support models improve the overall results, great. But in my own workflow, smaller models have mostly meant more corrections, so I typically just go straight to a larger model working alone. What I'm really looking for is a more capable main model. I hope Astra 6.1 launches alongside Terra

1 0 1 29 0

View Details

Diego Aud @dieaud91

14 hours ago

@ajs6888 Yeah that's exactly how I'm judging them. In my work, Sol-class models still tend to make mistakes that cost me time, while Astra understands what I want much more reliably and needs far fewer corrections. That's why I prefer frontier models.

0 0 0 47 0

View Details

Diego Aud @dieaud91

a day ago

@hakmgpt It looks like they still don't get it... 💀

The Information @theinformation

2 days ago

OpenAI's retreat from expensive side projects may be reversing as the company explores a deal to revive Sora while expanding its efforts in robotics, creator tools and personal devices. Read more details: thein.fo/47l0mJR

31 25 348 275K 92

0 0 0 45 0

View Details

Diego Aud @dieaud91

2 days ago

@rudi_bakken That would be a banger, but I'm keeping my expectations low so I don't get disappointed... 🥲

1 0 1 22 0

View Details

Diego Aud @dieaud91

2 days ago

@brksbrksbrks I suspect Ant is a bit ahead in terms of what it can deploy in the near term. Fable 5.5 looks solid based on the (credible) leaks, and I don't think OpenAI has a new version of Astra ready to deploy anytime soon... 💀

0 0 1 60 0

View Details

Diego Aud @dieaud91

2 days ago

That's unfortunate. Personally, I couldn't care less about image models. I'm only hyped for a new frontier model.

Lumina @LuminaBench

2 days ago

The new ChatGPT "aurora-red-alpha" model seems to be an image model: I just got it in arena

36 12 344 54K 30

1 0 17 864 1

View Details

Diego Aud @dieaud91

2 days ago

@brksbrksbrks Yeah. Releasing a sub-frontier image model instead of a frontier AI model would be even worse. I think OpenAI should stop trying to do everything and focus on just a few things, especially frontier models.

1 0 2 39 0

View Details

Diego Aud @dieaud91

4 days ago

@MattiaMerenda2 As an Italian, I share your frustration. Reading that post and hearing similar opinions made me feel sad for my country. That take is so shortsighted it's embarrassing coming from a grown man.

1 0 2 20 0

View Details

Diego Aud @dieaud91

5 days ago

@deredleritt3r I think that if they don't find a way forward on safety, this delay could end up costing them dearly.

0 0 4 351 1

View Details

Diego Aud @dieaud91

5 days ago

@scottjla Fair point, but it's still hard to know where to draw the line... The minority here on X tends to shape the decisions the company makes for the majority of its users

1 0 0 185 0

View Details

Diego Aud @dieaud91

6 days ago

OpenAI really screwed up with Astra 6.1. At this stage of the race, you can't afford to scrap a release if you're competing at the frontier. I think Astra 6.1 was meant to compete with Fable 5.5. If OpenAI has nothing to counter it with and expects to get away with its current models for more than 5-6 weeks (and based on the alleged leaks, I expect Fable 5.5 within the next 2 max 3 weeks), it's in for a rough time.

Token Gremlin @TokenGremlin

6 days ago

Bold take: Anthropic is becoming untouchable for OpenAI. Based on what's publicly available, I'd say Anthropic is 3 to 6 months ahead, and I don't think OpenAI will manage to close that gap anytime soon. My prediction? OpenAI stays in second place for at least another year.

106 11 508 93K 34

11 3 97 38K 15

View Details

Diego Aud @dieaud91

5 days ago

Good question. What I'm wondering is whether the issues that got Astra 6.1's release scrapped also prevent them from using it internally. If so, and unless they have another model ready, they might be one generation behind in terms of the models they can actually use over the next few weeks.

0 0 0 229 0

View Details

Diego Aud @dieaud91

5 days ago

@AndreTI Something tells me that as we move forward, generational leaps will create bigger discontinuities, and the advantage of being one generation ahead could compound faster and faster

1 0 1 298 0

View Details

Diego Aud @dieaud91

7 days ago

First output from Gemini 4 Argon looks solid and feels well-reasoned. I asked for its take on some generic, questionable nutritional advice given by a specialist to a patient of mine: it elegantly explained what works and what doesn't, while also breaking down the physiological mechanisms involved. If you're a medical specialist still giving outdated or imprecise nutritional advice, remember that your patients are just €20 away from much better guidance.