đ¤ w/149 - Capitalism Beats Safetyism At OpenAI
This week's article covers the reappointment of Sam Altman as CEO of OpenAI, highlighting the conflict between the for-profit and non-profit sides of the organisation, and the impact on AI safety
w/Wiser! #149 - 26th November 2023
Good Sunday Wiseristas!
Whatâs that saying about a weekâs a long time in politics? Well, a week on from last weekâs Wiser! about the unexpected sacking of Sam Altman, the boomerang CEO of OpenAI has returned.
Basically, here's the deal. Late Tuesday, OpenAI announced that Sam Altman was being brought back after a five-day campaign to have him reinstated as the boss. Remember, that in this five day period between firing and then reinstating him, OpenAI had appointed two other CEOs and Microsoft had announced he was now working for them!
The boomerang campaign was a combination of efforts by Altman himself, his allies, investors, and the employees, who had threatened to quit en mass and go work at Microsoft.
It was also clear that there was no smoking gun to support the boardâs decision to fire Altman. The Wall Street Journal reported, âthe board said that Altman had been so deft (at not telling the truth), they couldn't even give a specific example, according to the people familiar with the executivesâ.
In other words, the OpenAI board thought Altman was so good at being bad, they can't even tell you what he did bad. This kind of paranoid and conspiratorial thinking in a board room will always lead to bad decisions.
Which explains why OpenAI are overhauling the company's board of directors. Of those who voted to fire Altman, only Adam DiAngelo is staying on. New people are joining to replace those departing. Brett Taylor, an ex-Facebook and ex-Salesforce executive is coming on to the board as the new chairman. Larry Summers, the former US Treasury Secretary, is also coming on to the board.
So, what does it all mean and what have we learnt?
At the heart of the dispute was the conflict between the For Profit and the Not For Profit sides of OpenAI. Particularly a dispute between Altman and former board member Helen Toner over AI Safety.
Toner represented the nonprofit side of OpenAI. She recently co.authored a research paper that was, in part, about AI safety. And the thing that she and her co-authors wrote was that OpenAI's rival, Anthropic, which is co-founded by a bunch of former OpenAI people, had built their product more safely than OpenAI had.
In Altmanâs mind this was a betrayal and not helpful in the pursuit of commercial value. But for Toner she was just her doing her job, which is to make sure that AI gets built in the safest manner possible. Her job is not to protect the reputation and commercialisation of OpenAI.
Remember, the mission of OpenAI and the profit motive were in direct conflict with each otherâŚit was only a matter of time before one would win out.
By all accounts, it was this schism that lead to Altmanâs firing a week ago. But in just five days, it was clear that the money beat safety and Altman was back and Toner was gone.
It remains to be seen if/how this will change the mission of OpenAI. It can be expected that Microsoft is going to have a bigger hand in the governance of OpenAI going forward. Up until now, theyâve been a kind of a passive investor, but with so much at stake for Microsoftâs strategy in AI, itâs hard to see how they wont take an active role in OpenAI now.
Hereâs The Thing â This is so reminiscent of when Frances Haugen, a former Facebook employee turned whistleblower, accused Facebook of âprioritising profits over public safetyâ in 2021. Big Tech just wasnât doing âthe right thingâ then and it feels like deja vue all over again!
The bigger picture here is that in the war of words between the capitalists and the safetyists, those who say âpush forward because the opportunities are too greatâ, versus the âslow down, we donât know what weâve unleashed hereâ brigade, the capitalists have won.
Itâs telling that the people who are now in charge of the board of OpenAI are the kind of seasoned deal makers and Silicon Valley insiders that you would expect to govern a for profit technology company. They have replaced the academics and ideologues on the board who had been the worriers that AI could become too powerful and bring about the end of humanity.
Meanwhile, Anthropic have remained relatively silent, a smart move as the dust settles. Instead, theyâve released Claude 2.1 to retake top spot as the most powerful large language model. Theyâve also been quick to market a âAnthropic is 50% safer than ChatGPTâ tag line. See more belowâŚ
w/Supercharge
Learn AI in 5 minutes a day
AI Tool Report will teach you how to save time and earn more with AI.
Join 400,000+ free daily readers for trending tools, productivity boosting prompts, the latest news, and more.
w/News
Whatâs Happened This Week?
â Anthropic slashes AI pricing amid rising competition
Anthropic has lowered pricing of its conversational AI model Claude 2.1 due to increased competition and the rising number of open source alternatives. Open source proliferation poses a future challenge for closed AI vendors, as it offers greater customisation potential and lower costs.
Competition in the conversational AI market is driving down prices and pushing innovation, creating opportunities for enterprise adoption. âLearn more: VentureBeat
â Googleâs Bard AI chatbot can now answer questions about YouTube videos
Googleâs Bard AI chatbot can now answer questions about YouTube videos, extending its functionality. The YouTube Extension for Bard previously allowed users to find specific videos but now enables specific questions about video content.
The official announcement on Bardâs new feature said: âWeâre taking the first steps in Bardâs ability to understand YouTube videos. For example, if youâre looking for videos on how to make olive oil cake, you can now also ask how many eggs the recipe in the first video requires. Weâve heard you want deeper engagement with YouTube videos. So weâre expanding the YouTube Extension to understand some video content so you can have a richer conversation with Bard about it.â â Learn more: TechCrunch
5 Snippets Of News
Bill Gates says a 3-day work week where 'machines can make all the food and stuff' isn't a bad idea: AI won't take your job, but it will "change it forever," Bill Gates says.
Scientists 3D print a robotic hand with human-like bones and tendons. As a layer is printed, an optical scan IDs flaws and corrects them in the next layer.
IMF says central bank digital currencies can replace cash: âThis is not the time to turn back.â
USA demands 4 billion USD for termination of Binance criminal proceedings. The world's largest crypto exchange and its CEO Changpeng Zhao ("CZ") are accused of bank fraud, sanctions violations and money laundering, among other things.
Creative AI platform Runway launched a new feature that simplifies animation within image-to-video generation. Called Motion Brush, the feature allows users to simply paint an area or subject within a still image and Runway ML will automatically animate it across video frames.
w/Productivity
Hal9 - Enterprise ChatGPT
Chat with your enterprise databases using secure generative AI and empower business users in your team to do their own data analyses in seconds. â Learn more
Constant Contact
From AI automations to content generation, Constant Contact offers a range of advanced features to help you create highly targeted, effective campaigns without the hassle of writer's block. Say goodbye to the frustration of content creation and hello to more time doing what you love with who you love. â Learn More
MacPaw
Introducing Macpaw: the simplest way to clean, optimise, and secure your computer. With a range of easy-to-use apps, Macpaw makes keeping your device running smooth and secure a breeze. Enjoy a faster, smoother, and more secure computer experience with Macpaw. â Learn More
FREE TO WISER! READERS: If youâre new to generative AI and want to know the basics of how to get started, what you can do with AI and sample prompts to get you goingâŚdownload your copy of The Beginnerâs Guide To ChatGPT for FREE!
đ Show Your Support For Wiser!
Thank you for reading Wiser! If you got value and would like to support what Iâm doing, do this:
Forward this email: send it to anyone you know whoâs interested in the tech economy.
Make a donation: go to BuyMeACoffee to make a donation in the form of a virtual cup of coffee, they only cost âŹ2 each. And who doesnât love a coffee, right?
Check out my website: Youâll find links to everything I do at rickhuckstep.com.