OpenAI Announces New o3 Models

Welcome to another edition of Horizon AI,

OpenAI saved the best for the last day of its 12-day “Shipmas” event, presenting the successor to their o1 “reasoning” model, o3.

Let’s jump right in!

Read Time: 4.5’ min

Here's what's new today in the Horizon AI

• Chart of the week: AI Is About to Hit a Data Wall

• OpenAI Announces o3 and o3 Mini Model

• Free Resources

• AI tools to check out

• Video of the week

TOGETHER WITH MASTERWORKS

Invest with the art investment platform with 23 profitable exits

How has the art investing platform Masterworks been able to realize an individual profit for investors with each of its 23 exits to date?

Here’s an example: an exited Banksy was offered to investors at $1.039 million and internally appraised at the same value after acquisition. As Banksy’s market took off, Masterworks received an offer of $1.5 million from a private collector, resulting in 32% net annualized return for investors in the offering.

Every artwork performs differently — but with 3 illustrative sales (that were held for 1+ year), Masterworks investors realized net annualized returns of 17.6%, 17.8%, and 21.5%.

Masterworks takes care of the heavy lifting: from buying the paintings, to storing them, to selling them for you (no art experience required).

Now, you can SKIP THE WAITLIST when you use our exclusive link to get started.

The content is not intended to provide legal, tax, or investment advice. Past performance is not indicative of future performance. Investing involves risk.

“Net Annualized Return” refers to the annualized internal rate of return, or IRR, net of all fees and costs, to holders of Class A shares from the primary offering, calculated from the final closing date of such offering to the date the sale is consummated. A more detailed breakdown of the Net Annualized Return calculation for each issuer can be found in the respective Form 1-U for each exit. The 3 median returns above represent the ones closest in percentage to the median of the 12 exits with holding periods over 1 year.

Masterworks internally appraises artworks that are held by entities administered by Masterworks Administrative Services on an ongoing basis, and obtains an independent review of appraisals by a third-party appraiser on an annual basis.Appraisals are prepared in accordance with the 2024-2025 Uniform Standards of Professional Appraisal Practice (“USPAP”) developed by the Appraisal Standards Board of the Appraisal Foundation, although it is noted that there are potential conflicts of interest given that some or all individual members of the appraisal committee are employees of Masterworks and Masterworks retains an ownership interest in the subject artworks as well as ownership of the Masterworks Platform.

Masterworks compiles historical data from public auctions to produce metrics that we believe can be helpful in measuring and analyzing historical trends in artist markets and the historical price appreciation of specific artworks.

See important Reg A disclosures at masterworks.com/cd.

Chart of the week

AI Is About to Hit a Data Wall

• According to recent research by Epoch AI, we're facing a critical moment in AI development. By 2028, AI models will need as much text data as the entire public internet contains.

• AI training data needs are doubling yearly, while internet content grows only 10% annually. This challenge pushes innovation in AI development methods.

Source

AI News

OPENAI

OpenAI Announces o3 and o3 Mini Model

The company introduced o3, a model family and the successor to the O1 “reasoning” model, which includes both o3 and o3-mini, a smaller version fine-tuned for specific tasks.

Details:

• The O3 model family represents a significant leap in AI capabilities, excelling especially in coding, math, and scientific reasoning.

• Why o3 and not o2? The company skipped the name “o2” due to potential trademark conflicts with British telecom provider O2.

• The o3 model sets new records across several technical benchmarks, achieving 96.7% accuracy in the AIME 2024 math competition and scoring 87.7% in scientific reasoning on GPQA Diamond, outperforming typical PhD-level experts.

• o3-mini delivers strong performance with reduced computational costs and offers adjustable reasoning effort settings—low, medium, and high—providing flexibility for a wide range of tasks.

• The company also introduced a new safety technique called deliberative alignment, which uses the models' reasoning capabilities to better identify and handle potentially unsafe prompts.

OpenAI is rolling out o3 cautiously, with both models available for public safety testing until January 10th, 2025. o3-mini will fully release in late January, with o3 following shortly after. The journey toward AGI continues.

Source

Resources

📚 Build anything with Windsurf AI

💡 Tips for building effective agents

👀 Best-of-N (BoN) Jailbreaking, a simple black-box algorithm that jailbreaks frontier AI systems across modalities.

 📞 Talk to Gemini on your phone

AI Tools to check out

✅ Backflip AI: AI-powered design platform that translates user inputs into high resolution, 3D-printable models.

💥 Tempo Labs: Visual editor for React, powered by AI.

🦾 Impakt: AI Coach. Track workouts, get real-time feedback, analyze performance, and share progress.

🦦 Otterly.AI: Content & brand monitoring for AI-powered search experiences.

🌊 Surf: Combine people and posts from Bluesky, Mastodon, Threads, Flipboard, YouTube and more to create an awesome custom feed for whatever you want.

Video of the week

Former Google CEO Eric Schmidt on AI, China and the future

This interview covers a wide range of topics, including the promise and perils of AI, the AI race between the US and China, and the role of government in AI development.

Some of the key points discussed in the video:

• Schmidt believes that AI has the potential to revolutionize science and solve some of the world's most pressing problems, such as climate change and disease.

• He also warns of the potential dangers of AI, such as cyberattacks and the development of dangerous biological weapons.

• The US and China are competing in AI development, with the US at risk of losing its lead if it doesn’t take stronger action.

An insightful look into how AI could shape the future, from a veteran tech executive.

That’s a wrap!

Thanks for sticking with us to the end! Let’s stay connected on LinkedIn and Twitter.

Not subscribed yet? Sign up here and send it to a colleague or friend!

See you in our next edition!

Gina 👩🏻‍💻