Between the Sea and the Sky

Not sometimes. Not until the next update. Always. They proved it with math.

Even with perfect training data and unlimited computing power, AI models will still confidently tell you things that are completely false. This isn't a bug they're working on. It's baked into how these systems work at a fundamental level.

And their own numbers are brutal. OpenAI's o1 reasoning model hallucinates 16% of the time. Their newer o3 model? 33%. Their newest o4-mini? 48%. Nearly half of what their most recent model tells you could be fabricated. The "smarter" models are actually getting worse at telling the truth.

Here's why it can't be fixed. Language models work by predicting the next word based on probability. When they hit something uncertain, they don't pause. They don't flag it. They guess. And they guess with complete confidence, because that's exactly what they were trained to do.

The researchers looked at the 10 biggest AI benchmarks used to measure how good these models are. 9 out of 10 give the same score for saying "I don't know" as for giving a completely wrong answer: zero points. The entire testing system literally punishes honesty and rewards guessing.

So the AI learned the optimal strategy: always guess. Never admit uncertainty. Sound confident even when you're making it up.

OpenAI's proposed fix? Have ChatGPT say "I don't know" when it's unsure. Their own math shows this would mean roughly 30% of your questions get no answer. Imagine asking ChatGPT something three times out of ten and getting "I'm not confident enough to respond." Users would leave overnight. So the fix exists, but it would kill the product.

This isn't just OpenAI's problem. DeepMind and Tsinghua University independently reached the same conclusion. Three of the world's top AI labs, working separately, all agree: this is permanent.

Dragon age: the Veilguard

The “Dark Solas Theme” came on shuffle so that was that.

A user asked if there are instances wherein Rook will be referred to by their last name. "Yes, there are places where we refer to Rook by their last name. It's very contextual - we don't want to awkwardly insert a reference in where it doesn't make sense - but it does happen, though it is going to be more common with some backgrounds than others."

Q. "Does Solas have a type of tea he doesn't mind as much as others? Or are all of them equally detestable?" A. "Orange zinger."

A user asked about the ages of the companions. "We're not getting into specifics right yet, but the companions cover a broad range of ages and experience. They're all pretty established in the world and in their expertise by the time you meet them, though. You are recruiting a team of experts after all."

You can have body tattoos!

John Epler: "Echoing Corinne, the last few weeks have been an absolute whirlwind but also incredibly rejuvenating. I'm so excited to share more with y'all."

A user stated "There are no restrictions on lineage/faction/class combinations other than dwarves cannot be mages". Corinne added "Correct. I've really been enjoying the different combinations."

Q. "Will the gender identity of a trans/non-binary Rook be acknowledged in game by other characters? (beyond just pronouns I mean)". A. "There are moments when you can disclose and discuss, however, disclosure remains the right of the individual so you have to make that choice for yourself." (!!)

Re: release date. ''Exact date has not yet been revealed. Fall 2024 is as much as we can say right now.''

''Lots of great cameos in the game, but I'm not going to spoil anything. You'll have to play to see who shows up for yourself.''

''We can't reveal our upcoming marketing beats or media presence just yet. Stay tuned.''

Q. ''Can companion AI be customized - ie. DAO/DA2 style tactics or even something like Aggressive -> Defensive'' A. ''Very different system this time around. In the skill trees you can affect their behaviors, like setting you up for combos, or auto-healing you when health gets low. Each Companion is a little different mechanically.''

Q. ''Will we get the option to give our Rook more fantastical hair colors, like purple, blue, pink?''. A. ''Yes''

Q. ''How about heterochromatic eyes?'' A. ''Yes''

Corinne: ''We're so excited too! As much as I've played, I absolutely cannot wait to do my first playthrough when the game is out. Also cannot wait to hear about all of your decisions and experiences.''

Q. ''Will I be able to make a short qunari and a tall dwarf or are there limits to that? Do the qunari have to be buff?'' A. ''There are limitations, of course... but yes, you can make a Dwarf that is tall or a Qunari that is short, relative to their lineage. We stan the skinny Qunari''

Corinne: ''Need to get back to it, all. But it's always a pleasure to check-in and see the great questions. We'll do this again soon :)''

[source: the official BioWare Discord]