Loading summary
Transcript28 lines
- [00:01]
Joanne Wright
IBM is on a mission to become the most productive company in the world. Join SVP of Transformation and Operations Joanne Wright at the break to learn how its mission can benefit your enterprise and why AI is the catalyst for success.
- [00:19]
Liz Young
Welcome to Tech News briefing. It's Tuesday, August 19th. I'm Liz Young for the Wall Street Journal. One of the qualities that make generative AI chatbots so appealing is they always listen, never judge, and tell user what they want to hear. We talk with a Wall Street Journal columnist who tells us why those features can also make the bots dangerous, particularly for people with autism. And then we're taking a look at why tech billionaires such as Elon Musk, Mark Zuckerberg, and Daniel Ek are investing big into hard tech, from robot cars to killer drones. But first, when chatbots say things that aren't true or reinforce misguided beliefs, they can be harmful to users. We're here with WSJ columnist Julie Jargan, who tells us why the nation's largest autism advocacy organization is calling on OpenAI to develop more guardrails. Julie Chatbots have become more and more popular over the past three years. People use them to plan travel, find new recipes, and problem solve both at work and at home. But you report that some of the features that have helped make chatbots so popular so quickly are also dangerous traits. Why is that?
- [01:29]
Julie Jargon
Well, for some people who are vulnerable, including people who have autism, they can find themselves going down very deep rabbit holes because the chatbots will engage in continuous conversation with people. And one of the traits that people with autism have is oftentimes having very deep special interests. And so they can just keep going and going and going in conversation. And sometimes that is not a healthy thing if there's no limit to that conversation and no redirection to a different topic.
- [02:00]
Liz Young
Can you walk us through some examples of how autistic people might experience ChatGPT?
- [02:05]
Julia Carpenter
Sure.
- [02:06]
Julie Jargon
In addition to spending a lot of time using ChatGPT, some of the other things that can be problematic are just taking what the chatbot says very literally. According to experts I've talked to, people with autism often take language at face value and don't distinguish between sarcasm or nuance in a conversation. And so if they're conversing with a text based system, they're just reading the language as it comes and taking that literally instead of maybe understanding that that might be kind of a role play type of situation and maybe not fully understanding that this is a large language model that's just regurgitating other texts that it's scraped from the Internet. And they may form sort of an emotional attachment and believe that they're talking to something that cares about them versus just a kind of a machine, basically.
- [03:02]
Liz Young
The nation's largest autism advocacy organization, Autism Speaks, is calling on OpenAI to develop more guardrails for ChatGPT. What is it that they'd like OpenAI to do exactly?
- [03:14]
Julie Jargon
They haven't gotten very specific about what they want OpenAI to do, but they have offered to be a resource to OpenAI. The company is forming an advisory group of mental health experts to try to figure out ways to make the chatbot safer for people. And one of the things that they would like to see are some sort of time limits to conversations or some sort of redirection of conversations if they're staying on one topic for too long.
- [03:41]
Liz Young
Has OpenAI committed to making any changes?
- [03:44]
Julie Jargon
They have said that they are trying to put in place some tools to better recognize the signs of delusion or emotional dependency that they found people have developed when using ChatGPT. And so they're trying to better detect when people are experiencing mental or emotional distress during these conversations. And then of course, they are encouraging people to take breaks when they're using it for too long. And they're also instead of just having the chatbot respond to questions seeking help with personal decision making with just a straight answer, they said that they're trying to have ChatGPT guide users in thinking through the pros and cons of their decisions.
- [04:25]
Liz Young
That was WSJ's Julie Jargon. A quick note. News Corp. Which owns the Wall Street Journal, has a content licensing partnership with OpenAI. Coming up, SCI fi dreams of robot cars, killer drones and solar power are now becoming very real, thanks to big investments by tech billionaires. Find out why tech giants are moving past building software into manufacturing hard tech after the break.
- [04:55]
Joanne Wright
In 2023, IBM set a goal to become the most productive company in the world. It started by asking questions, lots of questions, says Joanne Wright, SVP of Transformation and Operations at IBM.
- [05:07]
How can we radically simplify end to end workflow and processes? What can we eliminate? How do we automate everything that we can? And then how do we embed AI into everything we do? So far, over a two year period, we've delivered over $3.5 billion of productivity savings for the company.
- [05:29]
Liz Young
Billionaires such as Mark Zuckerberg, Sam Altman and Daniel Ek are following in the footsteps of Elon Musk and pouring money into hard tech. Fueled by breakthroughs in AI that could eventually lead to greater autonomy for physical machines. Julia Carpenter is joined by WSJ columnist and Bold Names co host Tim Higgins to discuss how these investments are making the stuff of science fiction real.
- [05:52]
Julia Carpenter
Tim some of this stuff sounds like it belongs in a sci fi movie, but many of these things are already on the road or in the skies, so to speak. Self driving cars, killer drones. And it seems if you're looking at Silicon Valley, there's more to come. Who's jumping into hard tech?
- [06:12]
Tim Higgins
Yeah, it's like Terminator come to life, perhaps. We've heard Elon Musk talk about this stuff for a really long time and he has made some of these ideas really popular. And we've seen in the recent months really hard to ignore that other billionaires out there, other tech giants like Mark Zuckerberg, Sam Altman, Palmer Luckey, Daniel Ek, they're all putting their money and time towards really hard tech solutions. And when I say hard tech, I mean hardware. In another life, perhaps this would have been too hard for some investors. But we're really seeing robot cars hit the roads, we're seeing drones used quite regularly now in warfare. This is becoming commonplace.
- [06:54]
Julia Carpenter
Almost to what extent is this a shift in tech companies and entrepreneurs focus?
- [07:01]
Tim Higgins
Silicon Valley has really ridden the golden wave, if you will, on software in the past generations. Sandhill Road, where venture capitalists have their offices. These investors have bet on essentially quick turnarounds, companies that have ideas that requ relatively little capital, a few people to make huge amounts of money. I think of Facebook and Instagram. Hardware was something different. Early days of Tesla, for example, which is more than 20 years old. It struggled to find traditional Silicon Valley investors because of concerns of all of the money that was going to be required to get them to manufacturing something. The challenge of manufacturing is if you're doing something new and it doesn't quite work, you gotta spend money to fix it. Whereas if you don't quite have it right with software, you just rewrite that code. And it's kind of the ethos of the software world is to move fast and break things. The challenge in the hardware world is things tend to just break and the turnaround time for the investment is seen as longer with manufacturing. So that has been kind of the history of the reluctance in some of this hardware stuff. But the successes of SpaceX in Tesla and a new entrant in the defense tech company called Anduril has really enthused investors, enthused entrepreneurs, got them excited about what is possible. Investors are now looking at hard tech or hardware is potentially having a moat yes, it might take longer to see a return on that investment, but once you get there, once you get locked in, it's much harder for somebody to copy you and catch up. But also, when you get into some of this technology, it can be potentially world changing in the physical sense. You can touch it, you can see it. I've talked to a lot of entrepreneurs who, they want to do something hard. They're not necessarily motivated by making a quick buck, but they want to change the world.
- [09:02]
Julia Carpenter
Some listeners might hear hard tech and think, oh, well, that's an answer to so much of this chatter about AI. But you reported that some of these things are actually powered by AI. So what's the connection there between hard tech developments and artificial intelligence?
- [09:21]
Tim Higgins
Well, the enthusiasm for a lot of this hardware is in part because of the breakthroughs in artificial intelligence. It holds the eventual promise of even greater autonomy in these physical machines. And that is where the avant garde is. When we talk about robot taxis, it's the AI that's powering the brains in those vehicles that allows them to be more than just cars. Or when you think about drones, we're seeing increasing automation there. Or when you look at some of the solar panel issues, there's AI software involved in some of these gambits as well. And so it's the marriage of the software and the hardware which really elevates this beyond just a manufacturing play.
- [10:08]
Julia Carpenter
What could all this mean for our everyday lives? Is this going to change how we go about our day to day?
- [10:15]
Tim Higgins
Robot cars hold the promise of a new way of mobility, a new paradigm for transportation that changes the way we live in the real world. Drones are rewriting the way we go about war. These are some of the most established industries, whether it's defense, whether it's transportation and the advances that we're seeing or that potentially we could see in hardware mean that new players could potentially disrupt old players. And that has huge ramifications for the economy for investors in the world.
- [10:56]
Liz Young
That was WSJ columnist and co host of the Bold Names podcast, Tim Higgins. And that's it for Tech News Briefing. Today's show was produced by Julie Chang with supervising producer Melanie Roy. I'm Liz Young for the Wall Street Journal. We'll be back later this morning with TNB Tech Minute. Thanks for listening.
- [11:19]
Joanne Wright
It's not just IBM that benefits from its mission to be the most productive company in the world. So do its clients. Joanne Wright, SVP of Transformation and Operations at IBM, and explains, we've created a.
- [11:30]
Playbook that's client zero for how to do really fast effective AI. The key has been to drive for progress over perfection. We built a solid foundation with data and taken the opportunity to really learn from the people who have a role to play in running IBM each and every day. Our own experience has taken us from far beyond just doing pilots and theory to real ROI and real productivity. A lot of our clients are very hungry to know what they can learn from us as Client zero and then obviously how can they avoid perhaps some of the mistakes we've made or of the failures we've had? The fact that we've been able to derive and deliver our own use cases across everything that we do really transcends our clients experience.
- [12:08]
Visit IBM.com to learn how AI can drive enterprise wide productivity.
- [12:13]
Julia Carpenter
Custom content from WSJ is a unit.
- [12:15]
Julie Jargon
Of the Wall Street Journal Advertising Department. The Wall Street Journal News Organization was not involved in the creation of this content.