Daybreak - Friday Round-up: Who pays when Grok fails, Venezuela's crisis is Reliance's win, and more

Episode Date: January 8, 2026

In this episode we fill you in on four standout stories from the past week.First, a quick look at how the Venezuela crisis is benefiting Reliance and ONGC;Next, why the quick fashion promise stands on shaky ground;Third, how AI has been filling up Indian shopping carts; And finally, why the discourse about Grok AI is starting to sound a little lopsided. Tune in.Read The Ken's story on quick fashion here.Daybreak is produced from the newsroom of The Ken, India’s first subscriber-only business news platform. Subscribe for more exclusive, deeply-reported, and analytical business stories.

Hi, this is Rohan Dharma Kumar. If you've heard any of the Ken's podcasts, you've probably heard me, my interruptions, my analogies, and my contrarian takes on most topics. And you might rightly be wondering why am I interrupting this episode too. It's for a special announcement. For the last few months, I and Sita Ramon Ganeshan, my colleague and the Ken's deputy editor, have been working on an ambitious new podcast. It's called Intermission.

We want to tell the secret sauce stories of India's greatest companies. Stories of how they were born, how they fought to survive, how they build their organizations and culture, how they manage to innovate and thrive over decades, and most importantly, how they're poised today. To do that, Sita and I have been reading books, poring over reports, going through financial statements, digging up archives, and talking to dozens of people. And if that wasn't enough, we also decided to throw in video into the mix. Yes, you heard that right. Intermission has also had to find its footing in the world of multi-camera shoots in professional studios, laborious editing, and extensive post-production.

Sita and I are still reeling from the intensity of our first studio recording. Intermission launches on March 23rd. To get alert, as soon as we release our first video. episode, please follow intermission on Spotify and Apple Podcasts or subscribe to the Ken's YouTube channel. You can find all of the links at the ken.com slash I am. With that, back to your episode. Hello and welcome to daybreak. This episode is a little bit different from our usual ones where we tell you one important single business story. Instead, Rachel and I are together in the studio today. Hi. So we'll start by quickly running through

some of the most important business and tech headlines from the week. And then we'll pick one story that's been hard to ignore. This week, it's GROC, X's AI chatbot. We know what is going on in Venezuela with the US military invasion, right? The South American country has the biggest oil reserves in the world, but barely 1% of it is actually being used. Maduro, the president, he was captured, and some very bizarre things have followed,

as I'm sure everyone knows, including Trump saying that, he next wants to buy Greenland. So today, as we are recording this, Trump has said that the US government is going to take over Venezuela's oil exports. And interestingly, not long after this invasion happened, or share prices of reliance industries actually here in India went up by over 1% and hit a one year high in early trading. OnGC also, their shares rose up by 1 to 2% around the same time.

Why did that happen? So, turns out, UNGC, Videsh, it's this other subsidiary of UNGC, through which it makes international investments. It is owed about $500 million in unpaid dividends from a Venezuelan oil project that stopped almost a decade ago because of US sanctions and also the economic chaos that was happening in the country. So if exports restart under US control, that money could actually come back. to ONGC.

And the reliance angle is a little different. It could actually regain access to crude oil from the country that kind of suits the kind of refineries that it has. So, yeah, geopolitics is strange how one country's crisis becomes a gain for so many. So, Rachel, what about you? Any interesting stories that caught your eye? Actually, today's story from the ken was quite interesting. It was also kind of in the same vein as the episode you did on

10-minute deliveries. And the story is titled, Decathlon Tests Fashion through Grocery Speed QuickCommerce. I thought it was quite an insightful piece on how quick fashion is picking up in India's metropolitan cities. And our colleague reporter Pranati writes about how this new area of competition for e-commerce and quick commerce companies has been burning more cash than making it. Right. So like I mentioned, Snigda also just covered a similar point in our episode from yesterday,

which you should listen to if you haven't already. where she makes a case for how the 10-minute delivery should die. Because what's happening here is that this sector is so crowded and so saturated, that speed is becoming one of those very few differentiating factors that platforms can actually have new promises on. But what's also happening is that this target audience doesn't seem to be, like, see the need for this urgency. So some of the users that Pranati spoke to in this article,

they actually say that this is not a service they are looking forward to like using on regular. And, you know, already the cracks are showing. Platforms need at least 600 orders every day just to break even. But current order rates are averaging at like 170. Oh, wow. That's not even anywhere close. Yeah. So if this is one of like the last few strategies that they have left to, you know, differentiate themselves in the market. And if it's not showing a lot of promise, then where do we go from here? Yeah. It's the same for quick food and grocery delivery as well, you know. This was the last differentiator, where next, you know? But anyway, moving on to the next story,

it is actually a new report by Boston Consulting Group, and I thought it was very interesting. It's basically saying that India has become the world's biggest user of generative AI when we decide what to buy online or, you know, online shopping through apps. And actually, it's not that surprising because most of our shopping here in this country happens on phones. lately, we're comparing prices all the time, you know, different apps. Browsing actually has become, we're spending a lot of time just browsing moving from one app to the other, right?

And I think AI actually just plugs into that behavior and speeds it up. So it's a very natural fit that's happening. But there is also a flip side because brands which will not feed good information into these AI systems will risk being ignored, right? And AI suggestions will actually steer us consumers towards the few big players that exist while, you know, the smaller ones might disappear. And obviously that will change the way that we shop, what we are buying, who wins the market actually. And also what about, I was thinking, what about if AI companies monetize this and then it becomes another ad thing? You know, you pay to be featured. So what does that mean for consumers?

It makes so much sense in India. specifically because we're so cost-conscious. We're always looking like for the best deal. And if we're relying on AI to find the best deal for us, but obviously now there's an inbuilt incentive for companies and AI to make this another monetizable thing. So we're now in looking for the best deals, we're losing out on the best deals.

Probably. All right. So now on to the next big story of the week, which is GROC. Okay, so Rachel, what has been going on with this X's AI chatbot? Right, so what happened is that in the last week of December, X introduced this feature called edit image, which basically allows me as a user on Twitter to ask GROC to edit an image that somebody else has uploaded. Any user. Any user.

Okay. So, and any kind of an image. And it's been catching attention over the last week because a lot of users on Twitter have started asking GROC, to edit images of women by asking Grog to undress them or put them in bikinis. And in a lot of upsetting cases, it's also been happening to children. So... Men, of course. Why should we say users?

Mostly must be men, right? In most cases it is, yes. So the problem here is obviously that, you know, these people have not consented to have their images edited this way. So it's non-consensual sexualized content, which in general is not, is heavily regulated on platforms like X. And of course, governments across the world,

including India, UK, China, have sent notices to X saying that they need to take down this kind of content and ensure that they put certain, you know, rules in place so that GROC doesn't do this again. So now it might seem like, you know, users were just abusing, a loophole in this LLM, right? But turns out that's not really the case. In fact, X also has this feature called

Spicy Mode built into GROC AI which allows you to, you know, generate images or videos that are sexually explicit or NSFW and includes nudity and highly suggestive imagery. So that's what's drawing, you know, major controversy here. So they've actually made this feature.

It's like a conscious strategic thing that's been done. Exactly. Right? Because, I mean, we all know how Musk talks about, you know, his relationship with tech. He wants it to, he wants to make sure that Grok is less restricted. And, you know, he has this whole anti-voke agenda where he wants X to be like the bastion of free speech and not, you know, regulated by leftist liberal logic. So this is kind of part of all of that. Yeah.

But what's really. strange is that, you know, with AI development, you know, these basic safety standards are kind of inbuilt, right? That's why you can't just go to any random AI platform and ask them for, you know, sexualizing images or like, you know, undressing somebody. But it seems like, especially because of this spicy feature that it has, it seems like it's a very strategically done thing, right, on X's part. Because, ultimately. ultimately sex sells, right? Yeah.

And these kind of things do join. There's more controversy to it. There's more attention. You know, just by virtue of that, they'll probably make a lot of money. In fact, they just raised $20 billion from a bunch of different investors. Including NVIDIA. After all this has happened. After, in the same week, from NVIDia and Cisco's investment arms, notably, and a few other ventures as well.

Right. So this hasn't really affected their standards. so to say, except of course from a regulatory angle where a lot of governments have been asking them to restrict and make sure that this kind of content is prevented. So in the name of free speech, basically you're not ensuring that you maintain basic safety standards on your platform. Also this anti-voke thing, right? Like on a free speech platform, why on earth would you have this, you know, anti-woke agenda? Because woke people are actually pretty loud. So I feel like you'd want to have them on your platform. If you want more speech there.

True. But if you want more engagement, then... Yeah, exactly. So it seems like a marketing strategy that's targeted to a very specific kind of audience, which is very male, very white, very supremacist. And it seems to be working in excesses. Yeah, because of the current...

Look at the current political climate right now. I mean, of course, it's like, yeah, you're right. It's very well thought out, you know, this whole Antivoke thing. Actually, this reminds me of this episode of this episode of daybreak that I did probably two years ago. And this was when Grog was still new and X had written this introductory blog post about it. And it said that Grog is designed to answer questions with a bit of wit and it has a rebellious streak.

So if you don't have a sense of humour, then please don't come here, use it. And actually, it went quite out of control because it ended up with Musk being very angry. and I think that is when Grog started changing. The early version of Grog was of course, like they said rebellious. It was quite a potty mouth also. And on many instances, it actually bashed Elon Musk himself. And then I remember there was this one bit, I think it was a tweet that it wrote. And it said, well, well, well, if it isn't the man who put the twit in Twitter,

the one and only Elon Musk. So you want me to roast you based on your pose? I'll do my best. But I'm not sure you can handle it, you delicate little flower. And then it's done something, something. And then it's like maybe it's your inability to understand basic human emotions or your lack of self-awareness.

Or maybe it's just because you're a giant man-child who cannot resist a shiny new toy. Oh, my God. So Musk got so angry that, and then actually another thing happened. This professor ran a political leaning test on Grog. and he found that at the time Grok was even more left-leaning than Chad GPD and Musk saw this

and then he immediately responded saying I'm going to fix this it's going to be politically neutral but look where we are now doesn't have much of a sense of humour if he's the butt of the joke then on the other side we went to the dark side

but Rachel I wanted to ask you I think something that we miss talking about is what has been X's response to this whole situation. Right. So one of the only statements

that Musk has put out right now is that anybody who uses GROC in this way will be treated the same way as they would if they were to

upload illegal content. The area is murky because they're not directly uploading illegal content and GROC is in an inbuilt app is inbuilt into X

and they're using that to create illegal or illegal content. So, now that's a statement that X is put out. But the thing is GROC shouldn't be able to create this content in the first place.

Because other LLMs, other image generation vehicles like Open AI and Gemini, they don't allow this, which is exactly why GROC is catching so much. FACC. So there was no apology from X? There was an apology from GROC,

which is surprising because GROC is not human. it's not a person. It's not... It's human error. It's not... It's human error.

And it's not the thing that's supposed to take accountability over here. Yeah. So, yeah. That's very interesting because most of the news headlines that I've been reading

is GROC apologizes for blah, blah, blah. Like, it's... They're putting it on GROC. GROC, but what is GROC? Who is running GROC? Who is teaching GROC how to be? Yeah, that's kind of strange. That even journalism.

organizations are not catching on to that, you know? Like, how can GROC apologize? They can't. Yeah. One of other, one other statement of must is also him saying that blaming GROC is like blaming a pen. Like, it's on users to use tools with their own discretion and stuff like that. But then again, like we mentioned, it's, they're not, even if they ask, even if they

decide to use GROC in this way, GROC should not be complying. Yeah. Yeah. Yeah. Actually, and also somebody said, right, that Grog apologizes makes listeners think that the bot owns the incident, but it's the company's leader. And also about the apology in itself, you know, how did it come? It's just a string of words put together by an AI chatbot.

And that is based on previous patterns of human speech or human writing. Yeah. So that's what it boils down to. It's not really an apology. There is no apology and there is no accountability either. I think as of now they've started taking down the content, but there has been no statement about whether they're going to start building certain guardrails on GROC now, which is unlikely to happen because if their USPs that GROC is less restricted and it makes NSFW content,

then they can't go against that now, can they? Right. And Rachel, just one last thing quickly before we wrap. wrap up, I wanted to ask you, what does something like this mean for women in country like India? Well, I think it's actually a pretty huge safety risk because in India we have what we call an honor culture where a lot of our standing reputation is tied to the way we appear, especially in terms of sexual purity, so to say. So if there are pictures of women in compromising positions out there, then you know questions start coming up

like what kind of a person are you and things like that it'll be on the women to prove that actually it's an AI generated image it's not really them and now it's become so real yeah it's hard to tell and women have been beaten up hurt killed for much less than this right and on that grim note this is the end of the episode thank you so much for tuning in

and if you have any thought any feedback for us. Please write to us at podcast at the ken.com. You can write Daybreak Friday as a subject line and then we know you're talking about this episode. Yeah, see you next week.

Daybreak is produced from the newsroom of the Ken India's first subscriber-focused business news platform. What you're listening to is just a small sample of a subscriber-only offerings and a full subscription offers daily, long-form feature stories, newsletters and a whole bunch of premium podcast. To subscribe, head to the Ken.com and click on the red subscribe button on the top of the website.

Today's episode was hosted and produced by my colleague Snitha Sharma and edited by Rath CN.