Cognitive warmup. Google hasn’t really talked about this, but there’s a reason why I habitually call the them “the most sensible AI company, by far”. Hardly would you ever see a Google exec get into the sort of conversations that OpenAI and Anthropic tend to, all for the sake of attention. Secondly, Google has been silently working on and building real-world use cases for AI that solves problems on a much broader scale. India included. Society, governance, inclusivity, those sort of issues, much beyond the air-conditioned boardrooms with investors discussing AI investments to supposedly improve bottom-lines. Google’s habit of silently delivering AI utility extends to a new tool you’ve probably never heard of—Pinpoint. It is rather useful at finding some structure if you chuck troves of data its way. Great for professionals, such as journalists and academics. NotebookLM? No, this is a bit different.
As you can keep a ‘collection’ to up to 200,000 source files, you’re good to go. It’ll transcribe audio and video, summarise handwritten notes, scan images, PDF files and make them searchable, summarise documents and add labels such as people, organisations and locations, depending on context of that collection. All of this, of course, leverages Gemini models. Chances are, you could chance upon an insight, a data point or an information that may have otherwise bypassed your otherwise very trained radar. For, we are only human.
A real value add—many news organisations have made public collections, which can be found in the Explore section, on important themes and moments in time. I’m particularly impressed by the speed of transcription, though much like the voice notes transcription on any Android phone or iPhone, it does stumble momentarily with spellings or accuracy when the accent gets a bit complicated (though Pinpoint is generally more accurate and has better timestamps). And already a fan of the takeaway summarisation from audio recordings, if you ask for those, alongside the cleaner transcription. The latter extends to any document too. For some reason, heavy file uploads such as an audio or video file tend to fail at times—keep tabs, if you’re doing multiple at the same time.
PREVIOUSLY, ON NEURAL DISPATCH
INNOVATION, OPPORTUNITY AND DIVIDE
By the time you read this, the first ever UN Global Dialogue on AI Governance would be underway in Geneva. The momentum, and intent, is towards building a global consensus concerning AI governance. Specifically, to ensure the powerful AI models are regulated, the AI divide doesn’t widen any further, and people focus remains at the core of any governance.
Do read: Innovation exists everywhere, opportunity does not: UN pushes for AI governance
While it is ultimately up to democratically elected governments to set public priorities and legislate to meet competing interests, it is important that companies, along with civil society and other stakeholders, are also involved to advise on the relevant trade-offs and possibilities,” says Dr Claire Melamed, vice president of AI and Digital Cooperation Strategy at the UN Foundation.
HT asked her about the role AI companies will play in this conversation, and at what point. “AI companies play a key role in developing and disseminating this technology. However, as the power of AI grows, increasingly the companies themselves are asking governments to step in and help to set guardrails. Sam Altman of OpenAI and Dario Amodei of Anthropic have both recently called for this,” she says.
Egriselda López who is Permanent Representative of El Salvador to the United Nations and Co-Chair of the Global Dialogue on AI Governance, believes that this conversation about AI governance and regulation comes at a critical juncture, considering the pace at which AI’s capability claims are evolving.
Examples of that include Anthropic’s Claude Fable 5 and Mythos 5 models which were restricted for a certain time, Google adding ‘computer use’ function to Gemini 3.5 Flash models to execute workflow tasks like a human operator would, and OpenAI previewing its upcoming GPT-5.6 line including the Sol, Terra and Luna models.
Lopez believes it is important to identify signs of an AI divide early on. “It’s about infrastructure, it’s about data, skills, financing, institutional capacity, but also an ability to influence the standards that will affect everyone. Innovation exists everywhere. But opportunity does not,” she says.
At the India AI Impact Summit this year, Prime Minister Narendra Modi noted an intention to reverse a geographical divide regarding AI’s balance of power. “For too long, the Global South has been treated merely as a market for technologies created elsewhere. With AI, we are changing the narrative. The Global South will not just use AI; we will shape it, build it, and lead it,” he said, addressing the summit.
Any conversation that builds a workable template and method to govern and regulate AI, is more than welcome. UN’s hope is the Geneva dialogue lays the groundwork for that, with an intent that next year’s dialogue expected in New York to strengthen the foundations.
THE GREAT CIRCLE OF AI
Ford Motor Company, most likely influenced by grandiose pitches by AI companies (they aren’t the only corporate to have bought snake oil in the past year) thought it was a good idea to replace human engineers with AI cameras and AI algorithms for the process of manufacturing and design checks. Apparently, that isn’t something AI can do every well (among MANY other things), as the car maker has now realised.
The result? Ford is hiring back the greybeards and the white coats—as many as 350 of them—to take back the process from the machines. This number also tells me that at least 350 engineers would have been let go by Ford in the past few months, because true to corporate greed templates, they wouldn’t hire back more than they fired (if at all, the 350 re-hirings would still be a lower number than actual fired engineers).
“Mistakenly, we thought that by just introducing artificial intelligence and adjusting the design requirements that we had, that that would produce a high-quality product,” Charles Poon, Ford’s VP of vehicle hardware engineering, is reported to have said. This was, as the AI systems were put in place, a time when Ford thought AI was greater than humans. CEO Jim Farley had previously said AI will displace white-collar workers at massive scale. It simply isn’t, not by your experience. No matter how much corporate greed would force them down this avoidable circular path back to putting humans in charge of things.
Mark my words: Ford Motor Company will get these engineers to retrain AI systems, a couple of years later again claim AI is great, fire hundreds if not thousands of humans, realise AI isn’t still all that great, rehire. The circle of AI, we’re destined to suffer.
