Data: Language Models are Few-Shot Learners; Table: Axios Visuals

While GPT-3 has earned ecstatic reviews from many experts for its capabilities, some critics have pointed out clear issues around bias.

Why it matters: As AI becomes more powerful and more integrated into daily life, it becomes even more important to root out the persistent problem of bias and fairness.

What's happening: Researchers at OpenAI noted in the paper introducing GPT-3 that "internet-trained models have internet-scale biases." A model trained on the internet like GPT-3 will share the biases of the internet, including stereotypes around gender, race and religion.

  • As the table above from the paper shows, females were more often described with appearance-associated adjectives, while males were more often described with adjectives that spanned a wider spectrum of descriptions.
  • The paper also found that GPT-3 associated different races with different degrees of sentiment, with Black ranking consistently low.

In a Twitter thread, Facebook AI head Jerome Pesenti raised concerns that GPT-3 can "easily output toxic language that propagates harmful biases."

  • OpenAI CEO Sam Altman responded that he shared those concerns, and he argued that part of the reason the nonprofit was starting off GPT-3 in a closed beta was to do safety reviews before it went fully live.
  • He noted that OpenAI had introduced a new toxicity filter that was on by default.
  • The original paper also found that GPT-3 seemed less prone to bias than earlier, smaller models, offering some preliminary hope that size could help minimize the problem.

What to watch: A system that can generate near-human quality writing could be used for misinformation, phishing and other hacking efforts. And while malicious humans already do all of those things, GPT-3 and future AI systems could effectively scale those efforts up.

The bottom line: If AI produces racist or sexist content, it's because the system learned it by watching us. That puts the onus on programmers to curb their creations.

Go deeper: Rooting out AI bias

Go deeper

Aug 5, 2020 - World

How new tech raises the risk of nuclear war

Illustration: Annelise Capossela/Axios

75 years after Hiroshima and Nagasaki, some experts believe the risk of the use of a nuclear weapon is as high now as it has been since the Cuban missile crisis.

The big picture: Nuclear war remains the single greatest present threat to humanity — and one that is poised to grow as emerging technologies, like much faster missiles, cyber warfare and artificial intelligence, upset an already precarious nuclear balance.

Joe Biden picks Kamala Harris as running mate

Photo: Scott Olson/Getty Images

Joe Biden announced Tuesday that he has chosen Sen. Kamala Harris (D-Calif.) as his running mate — the first Black woman to be named to a major-party U.S. presidential ticket, and potentially the first woman vice president if Biden defeats President Trump.

The big picture: Harris was probably the safest choice Biden could have made among his running mate finalists. She has a national profile and experience with elected office, was vetted and tested in the Democratic presidential primaries and can boost Biden's fundraising.

Updated 1 hour ago - Politics & Policy

Coronavirus dashboard

Illustration: Annelise Capossela/Axios

  1. Global: Total confirmed cases as of 3 p.m. ET: 20,158,258 — Total deaths: 738,063 — Total recoveries: 12,388,686Map.
  2. U.S.: Total confirmed cases as of 3 p.m. ET: 5,116,791 — Total deaths: 164,137 — Total recoveries: 1,670,755 — Total tests: 62,513,174Map.
  3. States: Florida reports another daily record for deaths State testing plans fall short of demand.
  4. Axios-Ipsos poll: 1 in 2 has a personal connection to COVID-19.
  5. Business: Moderna reveals it may not hold patent rights for vaccine.
  6. 🏈 Sports: Big Ten scraps fall football season due to coronavirus.