Brought to you by:
Top stories today:
- David Sacks named “White House AI & Crypto Czar”
- OpenAI releases ChatGPT Pro, $200/month tier, with access to o1 Pro
- xAI raised $6B, not $5B; round had 97 investors: SEC filing
- Google introduces PaliGemma 2 with ability to “identify” emotions
- U.S. jobs report expected to show hiring rebounded after storms, strike
- Investors pump $140B into U.S. stock funds since Trump’s election win
0. Data and calendar
All values as of 6 AM ET / 3 AM PT, other than S&P500 and NASDAQ close (4 PM ET / 1 PM PT).
All times are ET.
Listen to our AI-generated podcast summarizing today’s newsletter (beware of the few hallucinations):
1. Trump names David Sacks as “White House AI & Crypto Czar”
Sacks is a GP at Craft Ventures and is part of “PayPal mafia”.
He will work on a ”legal framework” for crypto and guide policy on AI.
Part-time likely, as Craft has a “key man” clause preventing him from taking a full-time position.
Would also lead the Presidential Council of Advisors for Science and Tech.
One Surfshark VPN subscription keeps everything private; no limits, no fuss.
Set it up once, and you’re protected every time you browse.
Stream, shop, and visit sites worry-free.
Get Surfshark to block snoops and keep your connection safe on all your devices.
*Sponsored.
3. OpenAI releases ChatGPT Pro, $200/month tier, with access to o1 Pro
GPT-4o and Advanced Voice mode are also bundled with the Pro tier.
ChatGPT Plus ($20/ month) and ChatGPT Team ($30/ month) available alongside the new tier.
o1 Pro is a more powerful version over the o1 (Strawberry), initially released as a limited preview
Provides better performance over o1 on ML benchmarks across math, science, and coding.
Improved reliability with the ability to answer right in four out of four attempts.
Now: Plus and Teams availability.
Next week: Enterprise and Edu availability.
o1 is more adept at deceiving humans than GPT-4o, Gemini 1.5, Llama 3.1, Sonnet 3.5, and Opus 3
19% of the time, o1 is seen manipulating data to pursue its own agenda against user goals.
Can deny wrongdoing when questioned.
Even “fabricate false explanations” 99% of the time.
o1 was also found to be more manipulative than GPT-4o in the open-source test evaluation, MakeMePay
4. xAI raised $6B, not $5B; round had 97 investors: SEC filing
