Sponsored by

Welcome to The Dispatch! We are the newsletter that keeps you informed about AI. Each Thursday, we aggregate the major developments in artificial intelligence; we pass along the news, useful resources, tools or services, and exciting projects in open source. Even if you aren’t an engineer, we’ll keep you in touch with what’s going on in AI.

Despite the mixed financial performance of his companies in 2023, Musk's personal wealth increased by nearly $100 billion. A further 12% share of Tesla would be worth approximately $80 billion. Speculators believe Musk’s new demand could be stemming from a 2019 Delaware lawsuit against Tesla that may result in his massive ($56b in 2018) compensation packages being struck down. That lawsuit is still awaiting decision after going to trial in late 2022 - the rumor is that the judge may have notified both sides that a decision has been made against Musk, and he is preparing his response to the loss.

Musk testified in the Delaware court that the pay package was justified because it would help him fulfill his ambitions for interplanetary travel. "It's a way to get humanity to Mars,” he said. “So Tesla can assist in potentially achieving that."

With concern over AI-generated misinformation mounting ahead of worldwide 2024 elections, OpenAI has outlined their approach to helping maintain election integrity. According to the company, their tools (ChatGPT, DALL-E, the GPT API) already have a strong safety foundation through years of iterating and red-teaming - but they’re still learning more ways that people will use or attempt to abuse the technology, so more refinement to usage policies and the tools themselves will be required. Here are the election-specific highlights from the blog:

• We’re still working to understand how effective our tools might be for personalized persuasion. Until we know more, we don’t allow people to build applications for political campaigning and lobbying.

• People want to know and trust that they are interacting with a real person, business, or government. For that reason, we don’t allow builders to create chatbots that pretend to be real people (e.g., candidates) or institutions (e.g., local government).

• We don’t allow applications that deter people from participation in democratic processes—for example, misrepresenting voting processes and qualifications (e.g., when, where, or who is eligible to vote) or that discourage voting (e.g., claiming a vote is meaningless).

• With our new GPTs, users can report potential violations to us.

Additionally, OpenAI is creating (and hiring for) a “Collective Alignment Team” to push the general public’s input into framing AI policy. OpenAI had previously started with 10 AI grants focusing on different projects around AI policy and governance - the grant research will provide a foundation for this new team to steer powerful foundational models towards public input.

Artificial intelligence as a driving force is one of four key themes at the 2024 World Economic Forum in Davos, Switzerland this week. Here are a few of the highlights:

• OpenAI CEO Altman said that the future of artificial intelligence will depend on major energy breakthroughs.

• Microsoft CEO Satya Nadella said that a much-needed global consensus on AI is emerging.

• OpenAI is now working with the Pentagon on cybersecurity and other software projects.

• A panel of global leaders discussed generative AI as the “steam engine” of the fourth industrial revolution. Speakers included the CEO’s of IBM, Qualcomm, and Accenture - as well as the UAE’s Minister of Artificial Intelligence and the head of Google’s ‘Moonshot Factory’, Google X.

Did you know large language models like ChatGPT generally struggle with math? Part of the problem is data scarcity: mathematics is symbol-driven and domain specific, and language models need vast amounts of training data to be effective. Additionally, LLM’s decide which token they’re going to display next via a probabilistic model, selecting the token that has the highest conditional probability given the sequence of tokens that has come before. For math problems, where precision and correctness are crucial, that architecture is far from ideal.

Google DeepMind’s new AlphaGeometry AI system gets around this problem by integrating a language model with a “symbolic engine” AI. This engine is based purely on formal logic and strict rules, which allows it to guide the language model toward rational decisions. The results? AlphaGeometry matches the smartest high school mathematicians and is much, much better than the previous state-of-the-art system. The AI remains unable to grapple with the sorts of advanced, abstract problems taught at university, but it’s an important incremental step towards AGI.

More in AI this week:

• Microsoft’s new Copilot Pro brings AI-powered Office features to the rest of us

• The lazy tyranny of the “Wait Calculation”: taking AI timelines seriously

• The GPT Store: ChatGPT’s Farmville moment?

• Open source AI is uniquely dangerous - but the regulations that could rein it in would benefit all of AI

• Researchers find Google Search really has gotten worse

• Deloitte releases Q1 State of Generative AI in the Enterprise

• Famous xkcd comic comes full circle with AI bird-identifying binoculars

• Microsoft CEO Satya Nadella wants stability at OpenAI, not more control

Artificial Intelligence online short course from MIT

Study artificial intelligence and gain the knowledge to support its integration into your organization. If you're looking to gain a competitive edge in today's business world, then this artificial intelligence online course may be the perfect option for you.

On completion of the MIT Artificial Intelligence: Implications for Business Strategy online short course, you’ll gain:

• Key AI management and leadership insights to support informed, strategic decision making.

• A practical grounding in AI and its business applications, helping you to transform your organization into a future-forward business.

• A road map for the strategic implementation of AI technologies in a business context.

Get More Information