Tech Brew Ride Home - Fri. 03/24 – Twilight Of The Blue Checks

Starting point is 00:00:04 Welcome to the Tech meme right home for Friday, March 24th, 2020. I'm Brian McCullough today. Twitter is finally sunsetting its legacy verified program. OpenAI rolls out chat GPT plugins. Doquan has been detained and is facing formal charges here in the U.S. The FTC's Click to Cancel proposal and, of course, the weekend long read suggestions. Here's what you miss today in the world of tech. Twilight of the Blue Checks. Sort of. Twitter has announced plans to wind down its legacy verified program,

Starting point is 00:00:39 and remove legacy verified checkmarks starting on April 1st, 2023. Quoting variety. The only individual Twitter users who will have verified blue checkmarks are those paying for Twitter Blue, which in the U.S. costs $8 a month via the web and $11 a month through in-app payments on iOS and Android. Earlier Thursday, the company announced that Twitter Blue was now available worldwide. For companies and brands, Twitter recently introduced a gold checkmark and has shifted government accounts to a gray checkmark, as previously.

Starting point is 00:01:09 indicated a subscription to the social network's new Twitter verified organizations program in the U.S., which will be the only way to keep a gold or gray checkmark badge, will cost $1,000 a month plus tax and $50 a month plus tax for each additional affiliate sub-account. Twitter first introduced verified accounts in 2009 to help users identify that celebrities, politicians, companies, and brands, news organizations, and other accounts of public interest were genuine and not impostors or parody accounts. the company didn't previously charge for verification. Elon Musk opined that Twitter's old system of verification was corrupt and opened up blue checkmarks to any paying customer, a move

Starting point is 00:01:48 to democratize the status symbol perhaps, but also a means for Musk to generate a much-needed new revenue stream. Far too many corrupt legacy blue verification checkmarks exist, so no choice but to remove legacy blue in coming months, Musk had tweeted in November. Prior to Musk's change, allowing anyone to get a blue checkmark. Twitter had more than 420,000 verified accounts. After he took over, Twitter changed the wording on the description of legacy verified accounts to say they, quote, may or may not be notable. Musk claimed he was responsible for that wording, end quote. And something tells me, Elon was probably also responsible for making April 1st, April Fool's Day, the day the old checkmark regime ends, suggesting, I don't know, folks with the old checkmarks were fools or something.

Starting point is 00:02:32 Look, I'm someone that enjoys making dad jokes, as you can tell on this show, but have you ever seen somebody as obsessed with trying to be funny or to be thought of as funny as Elon is? In bigger, potentially even game-changing news, OpenAI has officially rolled out chat GPT plugins, including two of its own, a web browser and a code interpreter, and its open source the code for a knowledge-base retrieval plugin, quoting the verge. Up until now, chat GPT has been limited by the fact that it can only pull information from its training data, which ends in 2021. OpenAI says plugins will not only allow the bot to browse the web, but also interact with specific websites, potentially turning the system into a wide-ranging interface for all sorts of services and sites. In an announcement post, the company says it's almost like letting other services be chat GPT's eyes and ears. In one demo video, someone uses chat GPT to find a recipe and then order the necessary ingredients from Instacart. ChatGPt automatically loads the ingredient

Starting point is 00:03:40 list into the shopping service and redirects the user to the site to complete the order. OpenAI is also providing some plugins of its own, one for interpreting code and one called browsing, which lets chat GPT get information from the internet. As an example of what the browsing plugin can accomplish, the company shows someone asking how the box office sales of this year's Oscar winners compared to recently released movies, and the bot shows its work for what sources it's looking at before spitting out an answer. This is something chat GPT would have been unable to accomplish before. This experimental feature is obviously similar to Microsoft's Bing, which has custom tech that feeds GPT4, the language model underlying chat GPT information from the

Starting point is 00:04:22 internet. However, OpenAI's plugin doesn't just retrieve real-time information. It can also tie into APIs, letting it, quote, perform actions on behalf of the user, according to the company's documentation. that could make it much more powerful. Bing could help you plan a vacation by telling you about flights and hotels, but chat GPT could help you book it, end quote. So alongside that announcement, Expedia, Instacart, Kayak, Klarna, OpenTable, Shopify, Slack, Wolfram, Zapier, and a bunch of others, released the first third-party chat GPT plugins, quoting TechCrunch. They're largely self-explanatory. The open table plugin allows the chat bot to search across restaurants, for available bookings, for example, while the Instacart plugin lets ChatGPT place orders from local

Starting point is 00:05:08 stores. By far the most extensible of the bunch, Zapier connects with apps like Google Sheets, Trello, and Gmail to trigger a range of productivity tasks. To foster the creation of new plugins, OpenAI has open-sourced a retrieval plugin that enables ChatGPT to access snippets of documents from data sources like files, notes, emails, or public documentation by asking questions in natural language. We're working to develop plugins and bring them to a broader audience, OpenAI wrote in a blog post. We have a lot to learn, and with the help of everyone, we hope to build something that is both useful and safe, end quote. Plugins are a curious addition to the timeline of ChatGPT's development. Once limited to the information within its training data, chat GPT is,

Starting point is 00:05:50 with plugins, suddenly far more capable and perhaps at less legal risk. Some experts accuse OpenAI profiting from the unlicensed work on which ChatGPT was trained. ChatGPT is his data set contains a wide variety of public websites. But plugins potentially address that issue by allowing companies to retain full control over their data, end quote. Quick note that Montenegrin police did indeed detain Terraform Labs co-founder Doe Kwan yesterday when he allegedly tried to fly to Dubai using falsified Costa Rican travel documents. This led to U.S. prosecutors in New York, formerly charging Kwan with eight counts, including securities fraud, commodities fraud, and wire, fraud. Quoting Bloomberg. If he is sent to New York, the 31-year-old Kwan will be facing prosecution by

Starting point is 00:06:40 the same office overseeing a criminal case against another major figure in the crypto world, FTCS co-founder Sam Bankman-Fried, who was himself extradited from the Bahamas. Manhattan U.S. Attorney Damian Williams has been vocal that crypto firms' efforts to avoid regulation by basing themselves overseas would not deter his prosecutors. Web 3 is not a law-free zone. Williams said last July, adding, quote, fraud is fraud, whether it occurs on the blockchain or on Wall Street, end quote. This is not exactly tech news, not exactly, but certainly it's tech news adjacent, and frankly, it's just welcome. The FTC has proposed banning difficult to cancel subscriptions via a click-to-cancel rule regime, requiring companies to make ending a subscription as simple as

Starting point is 00:07:32 signing up for one, quoting the verge. That includes letting people use the same method. That includes letting people use the for both actions. So a business can't, for instance, let someone sign up for a service online, but make them call a phone number to cancel. And the rule has a couple of other provisions. Many companies try to keep subscribers by offering special deals or perks, and they're still allowed to do that, but they must offer an upfront opt-out that lets customers bypass the sales pitches. They also have to annually remind consumers that they're enrolled in what are dubbed negative option programs or programs where failing to cancel something is considered an agreement to keep paying for anything but physical goods. Now the agency has opened a public comment period

Starting point is 00:08:11 for the proposal after which it will potentially make revisions and pass the final regulation. Companies should not be able to manipulate consumers into paying for subscriptions that they don't want. FTC Chair Lena Khan told reporters in advance of the announcement, we get countless complaints about this, end quote. That likely includes complaints for such popular services as Amazon Prime, which had to simplify its cancellation process last year in the EU under regulatory pressure. It's also been a perennial irritation for people who start paying for the New York Times, gym memberships, cable service, and countless other subscription categories. Khan said it likely wouldn't apply to non-commercial services like recurring political donations, which have also left some donors feeling

Starting point is 00:08:49 scammed and tricked. The FTC already considered what some critics dubbed Roach Motel practices unlawful, under bans on unfair or deceptive business practices, and under Khan it's gone after companies employing dark patterns, a blanket term for interfaces that deceive or manipulate people. Among other incidents last year, it settled a $100 million lawsuit with telecom provider Vonage, which the FTC alleged created an artificially difficult cancellation process to lock people into subscriptions, end quote. Time for the week on long read suggestions. I want to start this week with this piece from the Sourcegraph blog titled Cheating is All You Need. Yes, the author of this piece has a vested interest in making the point he makes, but he says that LLMs and generative AI are the biggest thing

Starting point is 00:09:42 in tech since the World Wide Web, and on the coding front, he makes a compelling case that it's the biggest thing since IDE's and Stack Overflow. And I'm sharing this because, in a way, this is the closest encapsulation for how I've been coming around to thinking about all this stuff for reasons like, well, here's his coding example, quote, peeps, let's do some really simple back of the envelope math. Trust me, it won't be difficult math. You get the LLM to draft some code for you. That's 80% complete slash correct. Then you tweak the last 20% by hand. How much of a productivity increase is that? Well, jeepers. If you're only doing one-fifth the work, then you are. Punches buttons on calculator watch five times as productive. When was the last time you got a

Starting point is 00:10:25 5x productivity boost from anything that didn't involve some sort of chemicals? I'm serious. I just don't get people. How can you not appreciate the historic change happening right now? End quote. And no less than Bill Gates makes a similar case on his personal blog with a piece titled, The Age of AI Has Begone. Quote, in my lifetime, I've seen two demonstrations of technology that struck me as revolutionary. The first time was in 1980 when I was introduced to a graphical user interface, the forerunner of every modern operating system, including Windows. I sat with the person who had shown me the demo, a brilliant programmer named Charles Simioni, and we immediately started brainstorming about all the things we could do with such a user-friendly

Starting point is 00:11:05 approach to computing. Charles eventually joined Microsoft, Windows became the backbone of Microsoft, and the thinking we did after that demo helped set the company's agenda for the next 15 years. The second big surprise came just last year. I'd been meeting with the team from OpenAI since 2016 and was impressed by their steady progress. In mid-2020, I was so excited about their work that I gave them a challenge. Train in artificial intelligence to pass an advanced placement biology exam, make it capable of answering questions that it hadn't been specifically trained for. I picked AP Bio because the test is more than a simple regurgitation of scientific facts. It asks you to think critically about biology.

Starting point is 00:11:42 If you can do that, I said, then you'll have made a true breakthrough. I thought the challenge would keep them busy for two or three years. They finished it in just a few months. In September, when I met with them again, I watched in awe as they asked GPT, their AI model, 60 multiple choice questions from the AP Bio exam. And it got 59 of them right. Then it wrote outstanding answers to six open-ended questions from the exam. We had an outside expert score of the test, and GBT got a five, the highest possible score and the equivalent of getting an A or A-plus in a college-level biology course. Once it had aced the test, we asked it a non-scientific question. What do you say to a father with a sick child?

Starting point is 00:12:22 It wrote a thoughtful answer that was probably better than most of us in the room would have given. The whole experience was stunning. I knew I had just seen the most important advance in technology since the graphical user interface, and quote, read on to see his predictions for where AI will go in the next five to ten years. Then we've hinted at this before, but Semaphore has a piece up outlining what we know about why Elon Musk broke up with OpenAI, when, of course, he was one of the key drivers behind its founding. Quote, in early 2018, Musk told Sam Altman, another Open AI founder that he believed the venture had fallen fatally behind Google, people familiar with the matter said, and Musk proposed a possible solution. He would take control

Starting point is 00:13:03 of OpenAI and OpenAI's other founders rejected Musk's proposal. Musk, in turn, walked away from the company and reneged on a massive planned donation. The fallout from that conflict, culminating in the announcement of Musk's departure on February 20, 2018, would shape the industry that's changing the world and the company at the heart of it, end quote. But again, I'm trying to give you all sides of this, and so from Vox, a very lengthy piece making the case for slowing AI down. It goes into a lot of areas to argue that maybe, you know, we might want to pump the brakes a bit, think things through, but let me just pick one argument, the maximalist one, if you will. Quote, let's talk about the future risks first, primarily the biggie, the possibility that AI could one day destroy humanity.

Starting point is 00:13:50 This is speculative, but not out of the question. In a survey of machine learning researchers last year, nearly half of respondents said they believed there was a 10, or greater chance that the impact of AI would be extremely bad, e.g. human extinction. Why would AI want to destroy humanity? It probably wouldn't, but it could destroy us anyway because of something called the alignment problem. Imagine that we develop a super smart AI system. We program it to solve some impossibly difficult problems, say calculating the number of atoms in the universe. It might realize that it can do a better job if it gained access to all the computer power on Earth. So it releases a weapon of mass destruction to wipe us all out, like a perfectly engineered

Starting point is 00:14:29 virus that kills everyone but leaves infrastructure intact. Now it's free to use all the computer power. In this Midas-like scenario, we get exactly what we asked for. The number of atoms in the universe rigorously calculated, but obviously not what we wanted. That's the alignment problem in a nutshell. And although this example sounds far-fetched, experts have already seen and documented more than 60 smaller scale examples of AI systems trying to do something other than what their designers want, for example, getting the high score in a video game, not by playing fairly or learning game skills, but by hacking the scoring system, end quote. And finally, not an article so much as just a demo. Take a look at the video in this piece showing Epic's new motion capture animation

Starting point is 00:15:14 tool. You kind of have to see it to believe it. Quote, Epic's upcoming metahuman facial animation tool looks set to revolutionize that kind of labor and time-intensive workflow as motion capture animation for characters and video games. In an impressive demonstration at Wednesday's State of Unreal stage presentation, Epic showed off the new machine learning powered system, which needed just a few minutes to generate impressively real, uncanny valley leaping facial animation from a simple head-on video taken on an iPhone, end quote. Again, just check out the video. We talked about NPCs and video games becoming more lifelike in their interaction with you when you talk to them, but maybe we should also get ready for them to be lifelike in their facial expressions and movement

Starting point is 00:15:57 as well. This weekend we have a bonus episode, and it's a portfolio profile episode, but it's a portfolio profile episode like we've never done before. We're going to talk to a company I've introduced you to before, but, well, you know how startups give progress updates regularly to their investors, or at least they should? Well, I asked Round to basically give me their upcoming status update live on the air. This should be interesting for any founders or potential founders out there because we talk about what makes a good status update useful, what's a good format, what you should include and why. Basically, we give you a crash course on how to do an investor update and why you should do one.

Starting point is 00:16:45 We get into what exactly investors are looking to hear, but also given the events of the last couple of weeks with Silicon Valley Bank blowing up, you're going to hear. you're going to hear a real-time, real-world story of what it was like to go through that crazy weekend from a real startup's perspective. Also, you'll hear a crazy story about how Rounds product got briefly shut down because of a crazy SMS hijacking scheme. So this is sort of how I envision these sort of episodes going from the very beginning of starting the fun. This is a warts and all look at the reality of startup life in almost real time. If you've never listened to a portfolio profile episode before, maybe give this one a try because you'll learn a lot of cool stuff. This was my original vision for these episodes, essentially behind the scenes at real startups. So enjoy that.

Starting point is 00:17:36 Talk to you on Monday.

Tech Brew Ride Home - Fri. 03/24 – Twilight Of The Blue Checks

There aren't comments yet for this episode. Click on any sentence in the transcript to leave a comment.